R_scripts_final
Set working directory
setwd("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles")Load required libraries
library(tidyverse)
library(phyloseq)
library(viridis)
library(stringr)
library(vegan)
library(RColorBrewer)
library(BiocManager)
#BiocManager::install("microbiome")
library(microbiome)
#library(microbiomeutilities)
library(ggplot2)
library(knitr)
library(ggpubr)
library(pheatmap)
library(MASS)
library(gplots)
library(grid)
library(cowplot)
library(ggpubr)
library(DESeq2)
library(plyr)
library(multcomp)
library(randomForest)
library(corrplot)
library(ggrepel)
library(Hmisc)
source("HighstatLibV13.R")
library(ggcorrplot)
library(plyr)
library(lattice)
library(VennDiagram)
library(ggsn)
library(sf)
library(psych)
library(usefun)Load metadata
metadata <-read.table("metadata.txt", sep="\t", header = T, row.names = 1, fill = 1, dec = ".", na.strings = "NA")
metadata$long <- as.numeric(as.character(metadata$long))
metadata$lat <- as.numeric(as.character(metadata$lat))
metadata$DNA_ng_µl <- as.numeric(gsub(",", ".", gsub("\\.", "", metadata$DNA_ng_µl)))
metadata$A260_280 <- as.numeric(gsub(",", ".", gsub("\\.", "", metadata$A260_280)))
metadata$M_Seqs_trimmed <- as.numeric(gsub(",", ".", gsub("\\.", "", metadata$M_Seqs_trimmed)))
#str(metadata)Load Metaxa2 results and create a phyloseq object
metaxa_genus <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaxa_genus.txt")
# Create OTU table
OTU_metaxa <- metaxa_genus[,-1]
# Match sample ID order with metadata file
match <- match(rownames(metadata), colnames(OTU_metaxa))
OTU_metaxa <- OTU_metaxa[,match]
all(colnames(OTU_metaxa) == rownames(metadata))## [1] TRUE# Create tax table
tax_table_metaxa <- data.frame(str_split_fixed(data.frame(metaxa_genus) [,1], ";", 6))
colnames(tax_table_metaxa) <- c("Domain", "Phylum", "Class", "Order", "Family", "Genus")
# Check if samples are in order
identical(rownames(metadata), colnames(OTU_metaxa))## [1] TRUE# Combine into phyloseq object
metaxa_PHY <- phyloseq(otu_table(OTU_metaxa, taxa_are_rows=TRUE),
tax_table(as.matrix(tax_table_metaxa)), sample_data(metadata))
# Exclude taxa "Unknown", "Unclassified", "Eukaryota", "Mitochondria", "Archaea", "Chloroplast"
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Unknown"))
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Unclassified"))
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Eukaryota"))
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Mitochondria"))
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Archaea"))
metaxa_PHY <- subset_taxa(metaxa_PHY, !Domain %in% c("Chloroplast"))
# Add SSU counts to metadata
metadata$SSU_counts <- sample_sums(metaxa_PHY)
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaxa_PHY = subset_samples(metaxa_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaxa_PHY <- subset_samples(metaxa_PHY, country != "UK")
## Exclude biological / technical replicates
metaxa_PHY <- subset_samples(metaxa_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaxa_PHY_stat <- subset_samples(metaxa_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
#BH <- data.frame(c("BH01", "BH02", "BH03", "BH05", "BH06", "BH07", "BH09", "BH27", "BH29", "BH30", "BH34A", "BH35", "BH36", #"BH37", "BH38", "BH39", "BH44", "BH46", "BH47", "BH48", "BH49", "BH50", "BH58", "BH60", "BH61"))
#colnames(BH) <- c("sample")
#random_BH <- sample_n(BH, 8)
#BFH <- data.frame(c("BFH1", "BFH10", "BFH11", "BFH12", "BFH13", "BFH14", "BFH15", "BFH16", "BFH17", "BFH18", "BFH19", #"BFH20", "BFH21", "BFH22", "BFH23", "BFH25", "BFH28", "BFH29", "BFH30", "BFH31", "BFH32", "BFH33", "BFH34", "BFH35", #"BFH36", "BFH37", "BFH38A", "BFH39", "BFH4", "BFH40", "BFH41", "BFH6", "BFH8", "BFH9"))
#colnames(BFH) <- c("sample")
#random_BFH <- sample_n(BFH, 8)
# Sample set 1
metaxa_PHY_stat_equal <- subset_samples(metaxa_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#metaxa_PHY_stat_equal <- subset_samples(metaxa_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#metaxa_PHY_stat_equal <- subset_samples(metaxa_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
metaxa_PHY_WA_stat <- subset_samples(metaxa_PHY, continent == "West Africa")
# Create phyloseq object with only West African samples
metaxa_PHY_WA_WW_stat <- subset_samples(metaxa_PHY_WA_stat, category == "WA hospital effluent")Load data for rpoB
HMM_RESULT_TABLE <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/HMM_RESULT_TABLE.txt", row.names=1)
HMM_RESULT_TABLE$SUM = rowSums(HMM_RESULT_TABLE[,c(2,3)])
# Reorder samples to match metadata and add to metadata
# SUM
match <- match(rownames(metadata), rownames(HMM_RESULT_TABLE))
rpoB_counts <- HMM_RESULT_TABLE[match,]
metadata$rpoB_counts <- rpoB_counts$SUM
# R1
# Reorder samples to match metadata
match <- match(rownames(metadata), rownames(HMM_RESULT_TABLE))
R1_rpoB_counts <- HMM_RESULT_TABLE[match,]
metadata$R1_rpoB_counts <- rpoB_counts$R1Load Metaphlan3 results and create a phyloseq object
OTU_metaphlan <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/mod_merged_abundance_table_species.txt", header=T)
# Make sure tax tabe is in order
#tax_table_metaphlan <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", quote="\"", comment.char="")
#identical(tax_table_metaphlan$V1, OTU_metaphlan$clade_name)
tax_table_metaphlan <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", header=FALSE, sep=";")
colnames(tax_table_metaphlan) <- c("Kingdom", "Phylum", "Class", "Order", "Family", "Genus", "Species")
# Remove "__"
tax_table_metaphlan <- apply(tax_table_metaphlan, 2, function(y) (gsub(".__", "", y)))
match <- match(rownames(metadata), colnames(OTU_metaphlan))
OTU_metaphlan <- OTU_metaphlan[,match]
all(rownames(metadata) == colnames(OTU_metaphlan))## [1] TRUE# Combine into phyloseq object
metaphlan_PHY <- phyloseq(otu_table(OTU_metaphlan, taxa_are_rows=TRUE),
tax_table(as.matrix(tax_table_metaphlan)), sample_data(metadata))
# Check that sums are ~100
#sample_sums(metaphlan_PHY)
# Exclude Viruses, Eukaryota & Archaea
metaphlan_PHY <- subset_taxa(metaphlan_PHY, Kingdom != "Viruses" & Kingdom != "Eukaryota" & Kingdom != "Archaea")
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaphlan_PHY = subset_samples(metaphlan_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaphlan_PHY <- subset_samples(metaphlan_PHY, country != "UK")
## Exclude biological / technical replicates
metaphlan_PHY <- subset_samples(metaphlan_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaphlan_PHY_stat <- subset_samples(metaphlan_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 1
metaphlan_PHY_stat_equal <- subset_samples(metaphlan_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#metaphlan_PHY_stat_equal <- subset_samples(metaphlan_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#metaphlan_PHY_stat_equal <- subset_samples(metaphlan_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
metaphlan_PHY_WA_stat <- subset_samples(metaphlan_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
metaphlan_PHY_WA_WW_stat <- subset_samples(metaphlan_PHY_WA_stat, category == "WA hospital effluent")Load ResFinder results and create a phyloseq object
OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_resfinder))
OTU_resfinder <- OTU_resfinder[,match]
all(colnames(OTU_resfinder) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match
match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Divide by ARG gene lengths
## Get the lengths in terminal
# seqkit fx2tab --length --name --header-line resfinder.fasta > resfinder_lengths.txt
resfinder_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/resfinder_lengths.txt", header=FALSE, comment.char="#")
all(rownames(clusters_tax_table_resfinder$Gene) == resfinder_lengths$V1)## [1] TRUEOTU_resfinder_length_norm <- OTU_resfinder/resfinder_lengths[, 2]
# Normalization with Metaxa2 SSU counts
OTU_resfinder_length_SSU_norm <- t(t(OTU_resfinder_length_norm)/metadata$SSU_counts) * 1540
all(rownames(metadata) == colnames(OTU_resfinder_length_SSU_norm))## [1] TRUEidentical(OTU_resfinder_length_norm[2025, 5]/metadata$SSU_counts[5], OTU_resfinder_length_SSU_norm[2025, 5])## [1] TRUEall(rownames(OTU_resfinder_length_norm) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Reload to get new row numbers
write.table(OTU_resfinder_length_SSU_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_length_SSU_norm.txt",
row.names=T, sep = "\t", col.names = T)
OTU_resfinder_length_SSU_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_length_SSU_norm.txt", row.names=NULL)
OTU_resfinder_length_SSU_norm$row.names<-NULL
# Reload to get new row numbers
write.table(clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=NULL)
# Combine to phyloseq object
resfinder_PHY <- phyloseq(otu_table(OTU_resfinder_length_SSU_norm, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(clusters_tax_table_resfinder)))
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
resfinder_PHY = subset_samples(resfinder_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
resfinder_PHY <- subset_samples(resfinder_PHY, country != "UK")
## Exclude biological / technical replicates
resfinder_PHY <- subset_samples(resfinder_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
resfinder_PHY_stat <- subset_samples(resfinder_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 1
resfinder_PHY_stat_equal <- subset_samples(resfinder_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#resfinder_PHY_stat_equal <- subset_samples(resfinder_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#resfinder_PHY_stat_equal <- subset_samples(resfinder_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
resfinder_PHY_WA_stat <- subset_samples(resfinder_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
resfinder_PHY_WA_WW_stat <- subset_samples(resfinder_PHY_WA_stat, category == "WA hospital effluent")Normalization with rpoB
OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_resfinder))
OTU_resfinder <- OTU_resfinder[,match]
all(colnames(OTU_resfinder) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Normalization with rpoB counts
OTU_resfinder_rpoB_norm <- t(t(OTU_resfinder)/metadata$rpoB_counts)
all(rownames(metadata) == colnames(OTU_resfinder_rpoB_norm))## [1] TRUEall(rownames(OTU_resfinder_rpoB_norm) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Reload to get new row numbers
write.table(OTU_resfinder_rpoB_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_rpoB_norm.txt",
row.names=T, sep = "\t", col.names = T)
OTU_resfinder_rpoB_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_rpoB_norm.txt", row.names=NULL)
OTU_resfinder_rpoB_norm$row.names<-NULL
# Reload to get new row numbers
write.table(clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=NULL)
# Combine to phyloseq object
resfinder_rpob_PHY <- phyloseq(otu_table(OTU_resfinder_rpoB_norm, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
resfinder_rpob_PHY = subset_samples(resfinder_rpob_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
resfinder_rpob_PHY <- subset_samples(resfinder_rpob_PHY, country != "UK")
## Exclude biological / technical replicates
resfinder_rpob_PHY <- subset_samples(resfinder_rpob_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
resfinder_rpob_PHY_stat <- subset_samples(resfinder_rpob_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 3
resfinder_rpob_PHY_stat_equal <- subset_samples(resfinder_rpob_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#resfinder_rpob_PHY_stat_equal <- subset_samples(resfinder_rpob_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#resfinder_rpob_PHY_stat_equal <- subset_samples(resfinder_rpob_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
resfinder_rpob_PHY_WA_stat <- subset_samples(resfinder_rpob_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
resfinder_rpob_PHY_WA_WW_stat <- subset_samples(resfinder_rpob_PHY_WA_stat, category == "WA hospital effluent")Load CARD results for the samples sequenced here
CARD_metadata <- data.frame(resfinder_PHY_stat@sam_data)
# Headers simplified & ";" changed into tab in excel
OTU_CARD <-as.matrix(read.table("mod_COUNT_TABLE.txt", header= T, check.names = F, row.names = 1))
match <- match(rownames(CARD_metadata), colnames(OTU_CARD))
OTU_CARD <- OTU_CARD[,match]
# Normalization with Metaxa2 SSU counts
all(colnames(OTU_CARD) == rownames(CARD_metadata))## [1] TRUEOTU_CARD_norm <- t(t(OTU_CARD)/CARD_metadata$SSU_counts)
all(rownames(CARD_metadata) == colnames(OTU_CARD_norm))## [1] TRUEidentical(OTU_CARD[106, 4]/CARD_metadata$SSU_counts[4], OTU_CARD_norm[106, 4])## [1] TRUE# Normalization with HMM rpoB counts
#OTU_CARD_norm_rpoB <- t(t(OTU_CARD)/CARD_metadata$rpoB_counts)
#all(rownames(CARD_metadata) == colnames(OTU_CARD_norm_rpoB))
# Tax table (Modified in excel)
tax_table_CARD <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_CARD.txt", header=FALSE, sep="\t")
colnames(tax_table_CARD) <- c("Gene", "Class")
all(tax_table_CARD$Gene == rownames(OTU_CARD_norm))## [1] TRUE# Modify
tax_table_CARD_renamed = tax_table_CARD
tax_table_CARD_renamed$Gene <- as.matrix(gsub(pattern = "_[A-Z].*", replacement = "", tax_table_CARD$Gene))
# Have a look at the hits with duplicate names
dups <-tax_table_CARD_renamed[duplicated(tax_table_CARD_renamed$Gene)|duplicated(tax_table_CARD_renamed$Gene),]
# They are not those ones of interest here so we will not spend time on renaming them
# Reload to get new row numbers
write.table(OTU_CARD_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_CARD_norm.txt",
row.names=T, sep = "\t", col.names = T)
OTU_CARD_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_CARD_norm.txt", row.names=NULL)
OTU_CARD_norm$row.names<-NULL
CARD_PHY_stat <- phyloseq(otu_table(OTU_CARD_norm, taxa_are_rows = TRUE), sample_data(CARD_metadata),
tax_table(as.matrix(tax_table_CARD_renamed)))Load MGE results
OTU_MGE <-as.matrix(read.table("cp_MGE_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_MGE))
OTU_MGE <- OTU_MGE[,match]
all(colnames(OTU_MGE) == rownames(metadata))## [1] TRUE# Tax table
MGE_tax_table_trim <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", header=FALSE)
colnames(MGE_tax_table_trim) <- c("Gene", "Element", "Class")
# Reorder tax_table to match
match <- match(rownames(OTU_MGE), MGE_tax_table_trim$Gene)
MGE_tax_table_trim <- MGE_tax_table_trim[match,]
all(rownames(OTU_MGE) == MGE_tax_table_trim$Gene)## [1] TRUEMGE_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_lengths.txt", header=FALSE, comment.char="#", check.names = F)
match <- match(rownames(OTU_MGE), MGE_lengths$V1)
MGE_lengths <- MGE_lengths[match,]
all(rownames(MGE_tax_table_trim$Gene) == MGE_lengths$V1)## [1] TRUEOTU_MGE_length_norm <- OTU_MGE/MGE_lengths[, 2]
# Normalization with Metaxa2 SSU counts
OTU_MGE_length_SSU_norm <- t(t(OTU_MGE_length_norm)/metadata$SSU_counts) * 1540
all(rownames(metadata) == colnames(OTU_MGE_length_SSU_norm))## [1] TRUEall(rownames(OTU_MGE_length_SSU_norm) == MGE_tax_table_trim$Gene)## [1] TRUE# Reload to get new row numbers
write.table(OTU_MGE_length_SSU_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_MGE_length_SSU_norm",
row.names=T, sep = "\t", col.names = T)
OTU_MGE_length_SSU_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_MGE_length_SSU_norm", row.names=NULL)
OTU_MGE_length_SSU_norm$row.names<-NULL
# Reload to get new row numbers
write.table(MGE_tax_table_trim, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", row.names=F, sep = "\t", col.names = T)
MGE_tax_table_trim <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", row.names=NULL)
# Combine to phyloseq object
MGE_PHY <- phyloseq(otu_table(OTU_MGE_length_SSU_norm, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(MGE_tax_table_trim)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
MGE_PHY = subset_samples(MGE_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
MGE_PHY <- subset_samples(MGE_PHY, country != "UK")
## Exclude biological / technical replicates
MGE_PHY <- subset_samples(MGE_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
MGE_PHY_stat <- subset_samples(MGE_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 1
MGE_PHY_stat_equal <- subset_samples(MGE_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#MGE_PHY_stat_equal <- subset_samples(MGE_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#MGE_PHY_stat_equal <- subset_samples(MGE_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
MGE_PHY_WA_stat <- subset_samples(MGE_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
MGE_PHY_WA_WW_stat <- subset_samples(MGE_PHY_WA_stat, category == "WA hospital effluent")
# Get class 1 integrons
MGE_PHY_int <- tax_glom(MGE_PHY, taxrank = "Class")
MGE_PHY_int <- subset_taxa(MGE_PHY_int, Class == "intI1")
MGE_PHY_int_stat <- tax_glom(MGE_PHY_stat, taxrank = "Class")
MGE_PHY_int_stat <- subset_taxa(MGE_PHY_int_stat, Class == "intI1")
MGE_PHY_int_stat_equal <- tax_glom(MGE_PHY_stat_equal, taxrank = "Class")
MGE_PHY_int_stat_equal <- subset_taxa(MGE_PHY_int_stat_equal, Class == "intI1")
MGE_PHY_qac <- tax_glom(MGE_PHY_stat, taxrank = "Class")
MGE_PHY_qac <- subset_taxa(MGE_PHY_qac, Class == "qacEdelta")Load Virulence genes
OTU_VFDB <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/VF_COUNT_TABLE", sep=";")
## Modify headers in terminal
#sed -n 's/\(;\).*/\1/p' VF_COUNT_TABLE > output_headers.txt
#sed -i 's/;//g' output_headers.txt
#sed -i '/^$/d' output_headers.txt
output_headers <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/output_headers.txt", quote="\"", comment.char="")
colnames(output_headers) <- c("Accession")
# Check
all(output_headers$V1 == OTU_VFDB$X)## [1] TRUE# Reorder according to metadata
match <- match(rownames(metadata), colnames(OTU_VFDB))
OTU_VFDB <- OTU_VFDB[,match]
all(colnames(OTU_VFDB) == rownames(metadata))## [1] TRUE# Normalization with Metaxa2 SSU counts
OTU_VFDB_SSU_norm <- t(t(OTU_VFDB)/metadata$SSU_counts)
all(rownames(metadata) == colnames(OTU_VFDB_SSU_norm))## [1] TRUE#all(rownames(OTU_VFDB_SSU_norm) == tax_table_VFDB$V1)
# Reload to get new row numbers
write.table(OTU_VFDB_SSU_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_VFDB_SSU_norm.txt",
row.names=T, sep = "\t", col.names = T)
OTU_VFDB_SSU_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_VFDB_SSU_norm.txt", row.names=NULL)
OTU_VFDB_SSU_norm$row.names<-NULL
## Split columns so that there are more columns
tax_table_VFDB <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/VF.txt", header=FALSE, sep = " ")
new_columns <- str_split_fixed(tax_table_VFDB$V2, ";", 4)
colnames(new_columns) <- c("Gene", "Description", "Class", "Reference")
tax_table_VFDB <- tax_table_VFDB[-2]
colnames(tax_table_VFDB) <- c("Accession")
tax_table_VFDB <- cbind(tax_table_VFDB, new_columns)
# order as in the genemat
match <- match(output_headers$Accession, tax_table_VFDB$Accession)
tax_table_VFDB <- tax_table_VFDB[match,]
identical(tax_table_VFDB$Accession, output_headers$Accession)## [1] TRUE# Reload to get new row numbers
write.table(tax_table_VFDB, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_VFDB.txt", row.names=F, sep = "\t", col.names = T)
tax_table_VFDB <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_VFDB.txt", row.names=NULL)
identical(tax_table_VFDB$Accession, output_headers$Accession)## [1] TRUEVFDB_PHY <- phyloseq(otu_table(OTU_VFDB_SSU_norm, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(tax_table_VFDB)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
VFDB_PHY = subset_samples(VFDB_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
VFDB_PHY <- subset_samples(VFDB_PHY, country != "UK")
## Exclude biological / technical replicates
VFDB_PHY <- subset_samples(VFDB_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
VFDB_PHY_stat <- subset_samples(VFDB_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 1
VFDB_PHY_stat_equal <- subset_samples(VFDB_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#VFDB_PHY_stat_equal <- subset_samples(VFDB_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#VFDB_PHY_stat_equal <- subset_samples(VFDB_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
VFDB_PHY_WA_stat <- subset_samples(VFDB_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
VFDB_PHY_WA_WW_stat <- subset_samples(VFDB_PHY_WA_stat, category == "WA hospital effluent")Load crAssphage results
crassphage_table <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/crassphage_table.txt", header = T, row.names = 1)
# Match the order according to metadata
match <- match(rownames(metadata), rownames(crassphage_table))
crassphage_table <- crassphage_table[match,]
identical(rownames(metadata), rownames(crassphage_table))## [1] TRUE# Drop columns
crassphage_table = subset(crassphage_table, select = -c(coverage, alias) )
identical(rownames(metadata), rownames(crassphage_table))## [1] TRUEcrassphage_table <- t(crassphage_table)
crassphage_table_norm <- t(t(crassphage_table)/metadata$SSU_counts)
identical(crassphage_table[1, 12]/metadata$SSU_counts[12], crassphage_table_norm[1, 12])## [1] TRUEwrite.table(crassphage_table_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/crassphage_table_norm.txt", row.names=T, sep = "\t", col.names = T)
crassphage_table_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/crassphage_table_norm.txt")
# Reload to get new row numbers
all(rownames(metadata) == colnames(crassphage_table_norm))## [1] TRUEwrite.table(crassphage_table_norm, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/crassphage_table_norm.txt",
row.names=T, sep = "\t", col.names = T)
crassphage_table_norm <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/crassphage_table_norm.txt", row.names=NULL)
crassphage_table_norm$row.names<-NULL
# Create phyloseq object
crass_PHY <- phyloseq(otu_table(crassphage_table_norm, taxa_are_rows = T), sample_data(metadata))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
crass_PHY = subset_samples(crass_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
crass_PHY <- subset_samples(crass_PHY, country != "UK")
## Exclude biological / technical replicates
crass_PHY <- subset_samples(crass_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
crass_PHY_stat <- subset_samples(crass_PHY, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Create phyloseq object with equal group for the statistical analysis
## Include 8 samples per country
# Sample set 1
#crass_PHY_stat_equal <- subset_samples(crass_PHY, alias == "BFH1" | alias == "BFH19" | alias == "BFH25" | alias == "BFH31" | alias == "BFH6" | alias == "BFH15" | alias == "BFH39" | alias == "BFH10" | alias == "BH02" | alias == "BH09" | alias == "BH27" | alias == "BH34A" | alias == "BH44" | alias == "BH58" | alias == "BH61" | alias == "BH47" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 2
#crass_PHY_stat_equal <- subset_samples(crass_PHY, alias == "BFH34" | alias == "BFH35" | alias == "BFH18" | alias == "BFH32" | alias == "BFH14" | alias == "BFH41" | alias == "BFH22" | alias == "BFH23" | alias == "BH48" | alias == "BH61" | alias == "BH30" | alias == "BH47" | alias == "BH46" | alias == "BH60" | alias == "BH39" | alias == "BH44" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Sample set 3
#crass_PHY_stat_equal <- subset_samples(crass_PHY, alias == "BFH20" | alias == "BFH8" | alias == "BFH30" | alias == "BFH16" | alias == "BFH33" | alias == "BFH25" | alias == "BFH10" | alias == "BFH37" | alias == "BH49" | alias == "BH60" | alias == "BH36" | alias == "BH46" | alias == "BH02" | alias == "BH30" | alias == "BH61" | alias == "BH37" | alias == "FH1" | alias == "FH2" | alias == "FH3" | alias == "FH4" | alias == "FH5" | alias == "FH6" | alias == "FH7" | alias == "FH9")
# Create phyloseq object with only West African samples
crass_PHY_WA_stat <- subset_samples(crass_PHY, continent == "West Africa")
# Create phyloseq object with only West African wastewaters
crass_PHY_WA_WW_stat <- subset_samples(crass_PHY_WA_stat, category == "WA hospital effluent")Modelling ARG abundance
Data exploration
df<-data.frame(ARG_SUM=sample_sums(resfinder_PHY_stat),
intI1_SUM=sample_sums(MGE_PHY_int_stat),
MGE_SUM=sample_sums(MGE_PHY_stat),
hospital_section=as.factor(sample_data(resfinder_PHY_stat)$hospital_section),
SSU_counts=as.factor(sample_data(resfinder_PHY_stat)$SSU_counts),
hospital=as.factor(sample_data(resfinder_PHY_stat)$hospital),
category=as.factor(sample_data(resfinder_PHY_stat)$category),
city=as.factor(sample_data(resfinder_PHY_stat)$city),
country=as.factor(sample_data(resfinder_PHY_stat)$country),
continent=as.factor(sample_data(resfinder_PHY_stat)$continent),
R1_rpoB_counts=as.factor(sample_data(resfinder_PHY_stat)$R1_rpoB_counts),
no_of_beds=as.factor(sample_data(resfinder_PHY_stat)$no_of_beds),
daily_patient_visits=as.factor(sample_data(resfinder_PHY_stat)$daily_patient_visits),
long=as.factor(sample_data(resfinder_PHY_stat)$long),
lat=as.factor(sample_data(resfinder_PHY_stat)$lat),
A260_280=as.numeric(sample_data(resfinder_PHY_stat)$A260_280),
DNA_ng_µl=as.numeric(sample_data(resfinder_PHY_stat)$DNA_ng_µl),
M_Seqs_trimmed=as.numeric(sample_data(resfinder_PHY_stat)$M_Seqs_trimmed))
df$SSU_counts <- as.character(df$SSU_counts)
df$SSU_counts <- as.numeric(df$SSU_counts)
df$R1_rpoB_counts <- as.character(df$R1_rpoB_counts)
df$R1_rpoB_counts <- as.numeric(df$R1_rpoB_counts)
df$daily_patient_visits <- as.character(df$daily_patient_visits)
df$daily_patient_visits <- as.numeric(df$daily_patient_visits)
df$no_of_beds <- as.character(df$no_of_beds)
df$no_of_beds <- as.numeric(df$no_of_beds)
df$long <- as.numeric(gsub(",", ".", gsub("\\.", "", df$long)))
df$lat <- as.numeric(gsub(",", ".", gsub("\\.", "", df$lat)))
#str(df)
# Spatial positions
xyplot(long ~ lat, col = 1, pch = 16, data = df, scales = list(y = list(log = 10), x = list(log = 10)), aspect = "fill")
xyplot(long ~ lat, aspect = "fill", col = 1, pch = 16, data = df,
panel = function(x, y, ...) {
panel.grid(h = -1, v = -1, col.line = grey(0.9))
panel.xyplot(x, y, ...)
panel.text(x, y, labels = df$hospital,
col = grey(0.5), cex = 0.7, pos = 3, offset = 1)
})
df_Benin <- df[ which(df$country=="Benin"), ]
df_Finland <- df[ which(df$country=="Finland"), ]
df_BF <- df[ which(df$country=="Burkina Faso"), ]
df_Ben_BF <- df[ which(df$country == "Burkina Faso" | df$country == "Benin"), ]
gps <- data.frame(df$hospital, df$long, df$lat)
mytheme<-list()
mytheme$par.xlab.text$fontfamily="Times"
mytheme$par.ylab.text$fontfamily="Times"
mytheme$par.xlab.text$cex=1.25
mytheme$par.ylab.text$cex=1.25
mytheme$axis.text$cex=1.3
mytheme$axis.text$fontfamily="Times"
mytheme$par.main.text$fontfamily="Times"
mytheme$par.main.text$cex=2.25
p1 <- xyplot(long ~ lat, aspect = "fill", col = "#B2182B", pch = 19, data = df_Benin, cex = 3.5, par.settings=mytheme,
ylab = "Longitude", xlab = "Latitude", main = "Benin",
panel = function(x, y, ...) {
panel.grid(h = -1, v = -1, col.line = grey(0.9))
panel.xyplot(x, y, ...)
panel.text(x, y, labels = df_Benin$hospital,
col = "white", cex = 1.5, position = "top", offset = 1, fontfamily = "Times")
})
p2 <- xyplot(long ~ lat, aspect = "fill", col = "#44AA99", pch = 19, data = df_BF, cex = 3.5, par.settings=mytheme,
ylab = "Longitude", xlab = "Latitude", main = "Burkina Faso",
panel = function(x, y, ...) {
panel.grid(h = -3, v = -1, col.line = grey(0.9))
panel.xyplot(x, y, ...)
panel.text(x, y, labels = df_BF$hospital,
col = "white", cex = 1.5, position = "top", offset = 1, fontfamily = "Times")
})
p3 <- xyplot(long ~ lat, aspect = "fill", col = "#2166AC", pch = 19, data = df_Finland, cex = 3.5, par.settings=mytheme,
ylab = "Longitude", xlab = "Latitude", main = "Finland",
panel = function(x, y, ...) {
panel.grid(h = -1, v = -1, col.line = grey(0.9))
panel.xyplot(x, y, ...)
panel.text(x, y, labels = df_Finland$hospital,
col = "white", cex = 1.5, position = "top", offset = 1, fontfamily = "Times")
})
compass <- north2(p1, symbol = 1)
A <- plot_grid(p1, p2, nrow = 2, rel_widths = c(1, 1))
B <- plot_grid(p3, NULL, compass, NULL, ncol = 1, rel_widths = c(1, 1, 0.333, 1), rel_heights = c(1, 0.333, 0.333, 0.333))
plot_grid(A,B)
#ggsave(filename = "map.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Outliers
MyVar <- c("country", "city", "hospital", "hospital_section", "SSU_counts", "R1_rpoB_counts", "continent", "no_of_beds", "daily_patient_visits", "ARG_SUM", "intI1_SUM", "A260_280", "DNA_ng_µl", "M_Seqs_trimmed")
Mydotplot(df[,MyVar])
# Identify the outlier
#par(mfrow = c(1, 1))
#plot(x = df$country,
# y = df$SSU_counts)
#identify(x = df$country,
# y = df$SSU_counts)
# Remove it?
#df <- df[-60, ]
#dim(df)
# No, we shoud not remove it since it is most likely a
# a real observation, and there are not enough observations to be sure.
# Collinearity
MySel <- c("ARG_SUM", "intI1_SUM", "country", "city", "hospital", "hospital_section", "SSU_counts", "R1_rpoB_counts", "continent", "A260_280", "DNA_ng_µl", "M_Seqs_trimmed")
Mypairs(df[,MySel])
# Obviously, there is collinearity between continent and country.
# We can continue with onlu country. Similarly we can apply only SSU_counts
# as they are colloinear with R1_rpoB_counts.
#par(mfrow = c(1, 2))
boxplot(ARG_SUM ~ country, data = df, main = "ARGs")
boxplot(intI1_SUM ~ country, data = df, main = "intI1")
# Numer of zeros in the response variable
#100 * sum(df$ARG_SUM == 0) / nrow(df)
#number of observations per level of a categorical covariate
table(df$country)
table(df$hospital)
# Unequal number of hospitals. Let's not include that as a covariate.
# Conditional boxplots to study the collinearity of the covariates
par(mfrow = c(2, 2))
boxplot(SSU_counts ~ country,
xlab = "country",
data = df,
main = "SSU_counts")
boxplot(R1_rpoB_counts ~ country,
xlab = "country",
data = df,
main = "rpoB counts")
boxplot(M_Seqs_trimmed ~ country,
xlab = "country",
data = df,
main = "M_Seqs_trimmed")
boxplot(DNA_ng_µl ~ country,
xlab = "country",
data = df,
main = "DNA_ng_µl")
# There seems to be variation in the SSU_counts
# between hospital WW collected in different countries.
# Although the SSU and rpoB counts seemed to be collinear,
# it seems that the SSU_counts have a greater variance.
# Let's pick that among these two.
# Interactions
# SSU counts and ARG abundances by country
p <- ggplot(data = df,
aes(x = SSU_counts,
y = ARG_SUM))
p <- p + geom_point()
p <- p + xlab("SSU_counts") + ylab("ARGs")
p <- p + theme(text = element_text(size=15))
p <- p + geom_smooth(method = "glm")
p <- p + facet_grid(~country)
p 
# There seems to be some kind of interaction defined by country,
# since the smoothers are poistive/negative/with different slopes
# among different countries.
p <- ggplot(data = df,
aes(x = SSU_counts,
y = intI1_SUM))
p <- p + geom_point()
p <- p + xlab("SSU_counts") + ylab("IntI1")
p <- p + theme(text = element_text(size=15))
p <- p + geom_smooth(method = "glm") #You can also do glm
p <- p + facet_grid(~country)
p
# There seems to be some kind of interaction defined by country.
# However, the quality of the data (e.g. only few observations in Finland)
# is too low to be able to include this interaction to the model.
# Maybe it is good not to include this interaction to the model due to the low quality data.
# The relationship of class 1 integron to the ARG sum
p <- ggplot()
p <- p + geom_point(data = df,
aes(y = ARG_SUM,
x = intI1_SUM),
shape = 1,
size = 2)
p <- p + xlab("intI") + ylab("args")
p <- p + theme(text = element_text(size=15))
p <- p + geom_smooth(data = df,
method = glm,
aes(x = intI1_SUM,
y = ARG_SUM))
p <- p + facet_grid(. ~ country)
p
# In Finland the steeper correlation relies solely on one datapoint
# Thus it is difficult to say whether this is real.
# Due to the low quality of this data, this interactions should not be included in the model.
# We can however build a model and test whether it is significant.
# Does the hospital section have a role?
par(mfrow = c(1, 2))
boxplot(ARG_SUM ~ hospital_section,
xlab = "country",
data = df,
main = "hospital_section")
boxplot(intI1_SUM ~ hospital_section,
xlab = "country",
data = df,
main = "hospital_section")
# No.
# ANALYSIS
# The model:
#Call: glm(formula = ARG_SUM ~ country + SSU_counts + intI1_SUM + country:intI1_SUM,
# family = Gamma(link = "log"), data = df)
# would have been ideal, but was discarded due to the fact that the interactions were not
# based on high quality data and relied on single influental data points.
# MODEL SELECTION
# Let's fit a model with Gamma distribution with a log link.
Mall <- glm(ARG_SUM ~ country + SSU_counts, + intI1_SUM,
data = df, family="Gamma"(link="log"))
summary(Mall)
M1 <- glm(ARG_SUM ~ country,
data = df, family="Gamma"(link="log"))
summary(M1)
step(M1, direction = "both", scope=formula(Mall))
# We are left with:
#Call: glm(formula = ARG_SUM ~ country + intI1_SUM + SSU_counts, family = Gamma(link = "log"),
# data = df)
# Let us build the new selected model
M1 <- glm(ARG_SUM ~ country + intI1_SUM + SSU_counts,
data = df, family="Gamma"(link="log"))
summary(M1)
# Estimate Std. Error t value Pr(>|t|)
#(Intercept) -7.639e-01 1.502e-01 -5.085 3.65e-06 ***
#countryBurkina Faso -3.405e-01 1.021e-01 -3.335 0.001443 **
#countryFinland -6.439e-01 1.637e-01 -3.935 0.000213 ***
#intI1_SUM 1.878e+00 2.570e-01 7.308 6.28e-10 ***
#SSU_counts 7.411e-06 2.196e-06 3.375 0.001278 **
# Generalized R^2:
(21.8860 - 7.9953) / 21.8860
# R^2 = 0.6346843
# Let's save also the simpliest model
M0 <- glm(ARG_SUM ~ country + SSU_counts, data = df, family= "Gamma"(link="log"))
summary(M0)
# Generalized R^2:
(21.886 - 14.580) / 21.886
# R^2 = 0.3338207
# MODEL VALIDATION
# Homogeneity
# Plot residuals vs fitted values
F1 <- fitted(M1)
E1 <- resid(M1, type = "pearson")
par(mfrow = c(1,1), cex.lab = 1.5, mar = c(5,5,2,2))
plot(x = F1,
y = E1,
xlab = "Fitted values",
ylab = "Pearson residuals")
abline(h = 0, lty = 2)
# No patterns, we are good.
# Influential observations
par(mfrow = c(1, 1))
plot(cooks.distance(M1), type = "h", ylim = c(0, 1))
abline(h = 1)
# There are no influental observations
#cooksd <- data.frame(cooks.distance(M1))
#rownames(cooksd) <- rownames(M1[["data"]])
#cooksd[cooksd$cooks.distance.M1. > 1,]
# Normality
par(cex.lab = 1.5, mar = c(5,5,2,2))
E1 <- resid(M1)
hist(E1, breaks = 15, xlab = "Residuals", main = "")
# Independence due to model misfit
df$E1 <- E1 #Put E1 inside the df
MySel <- c("SSU_counts", "intI1_SUM", "country", "M_Seqs_trimmed")
MyMultipanel.ggp2(Z = df,
varx = MySel,
vary = "E1",
ylab = "Residuals",
addSmoother = TRUE,
addRegressionLine = FALSE,
addHorizontalLine = TRUE) 
# Some / No clear non-linear patterns in these graphs.
boxplot(E1 ~ country,
data = df,
ylab = "Residuals")
abline(h = 0)
# Looks good.
# Check for spatial dependency
# Spatial patterns in the residuals?
MyCex <- 3 * abs(E1) / max(E1)
MyCol <- ifelse(E1 > 0, "red", "blue")
xyplot(long ~ lat, data = df, cex = MyCex, col = MyCol)
xyplot(long ~ lat, data = df_Benin, cex = MyCex, col = MyCol)
xyplot(long ~ lat, data = df_BF, cex = MyCex, col = MyCol)
xyplot(long ~ lat, data = df_Finland, cex = MyCex, col = MyCol)
# In general, some spatial dependency can be detected.
# However, the number of samples in different locations is quite low.
# MODEL INTERPRETATION
M1 <- glm(ARG_SUM ~ country + SSU_counts + intI1_SUM,
data = df, family="Gamma"(link="log"))
summary(M1)
M0 <- glm(ARG_SUM ~ country,
data = df, family="Gamma"(link="log"))
summary(M0)
# Fit model
# Predictions
glht.M1 <- glht(M1, mcp(country = "Tukey"))
summary(glht(glht.M1))
pvalues <- tibble::tribble(
~group1, ~group2, ~p,
"Burkina Faso", "Finland", 0.07249,
"Benin", "Burkina Faso", 0.00214,
"Benin", "Finland", 0.001,
)
pvalues
cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
# Plot
dfA <- cbind(df, Mean = predict(M1, newdata = df, type = "response"), SE = predict(M1,
newdata = df, type = "response", se.fit = T)$se.fit)
resfinder_M1 <- ggplot(dfA, aes(x = country, y = Mean)) + scale_color_manual(values=cols) +
geom_line() +
geom_jitter(data = dfA, aes(x = country, y = ARG_SUM, color = country), size = 12, alpha = 1, width = 0.3) +
# geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.5, lwd = 0.75) + geom_point(size = 0.9) +
theme_linedraw() +
theme(axis.text.x = element_text(angle = 0, size = 32, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 28, family = "Times"),
axis.title.y = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
legend.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 40, family = "Times")) +
labs(y = "ARGs normalized to 16S rRNA", x = "") +
guides(color = FALSE, alpha = FALSE) +
labs(title = "Sum relative abundance of ARGs (ResFinder)")
resfinder_M1 + stat_pvalue_manual(pvalues, label = "p", y.position = 4, step.increase = 0.05, tip.length = 0.01, size = 9)
#ggsave(filename = "resfinder_sum_M1_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# OR
MyData <- ddply(df,
.(country),
summarise,
intI1_SUM = seq(min(intI1_SUM),
max(intI1_SUM),
length = 25),
SSU_counts = seq(min(SSU_counts),
max(SSU_counts),
length = 25))
head(MyData)
P1 <- predict(M1, newdata = MyData, se = TRUE, type = "response")
MyData$Pred <- P1$fit #Predicted values
MyData$selo <- P1$fit - 1.96 * P1$se.fit #lower bound
MyData$seup <- P1$fit + 1.96 * P1$se.fit #lower bound
head(MyData)
# Plot
#cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
#p <- ggplot()
#p <- p + geom_point(data = df, aes(y = ARG_SUM, x = intI1_SUM, col = country), shape = 16, size = 3)
#p <- p + xlab("intI1_SUM") + ylab("ARG_SUM")
#p <- p + theme(text = element_text(size=15))
#p <- p + geom_line(data = MyData, aes(x = intI1_SUM, y = Pred, col = country, group = country))
#p <- p + geom_ribbon(data = MyData, aes(x = intI1_SUM, ymax = seup, ymin = selo, group = country,col = country, fill = country),
# alpha = 0.5) + scale_color_manual(values=cols) + theme_minimal()
#ggsave(filename = "predict_M1_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Simple model
glht.M0 <- glht(M0, mcp(country = "Tukey"))
summary(glht(glht.M0))## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored
## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored##
## Simultaneous Tests for General Linear Hypotheses
##
## Linear Hypotheses:
## Estimate Std. Error z value Pr(>|z|)
## Burkina Faso - Benin == 0 -0.5189 0.1181 -4.394 <0.001 ***
## Finland - Benin == 0 -1.0266 0.1821 -5.639 <0.001 ***
## Finland - Burkina Faso == 0 -0.5078 0.1761 -2.883 0.0102 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)pvalues <- tibble::tribble(
~group1, ~group2, ~p,
"Benin", "Burkina Faso", 0.001,
"Benin", "Finland", 0.001,
"Burkina Faso", "Finland", 0.0105
)
pvalues## # A tibble: 3 x 3
## group1 group2 p
## <chr> <chr> <dbl>
## 1 Benin Burkina Faso 0.001
## 2 Benin Finland 0.001
## 3 Burkina Faso Finland 0.0105dfA <- cbind(df, Mean = predict(M0, newdata = df, type = "response"), SE = predict(M0,
newdata = df, type = "response", se.fit = T)$se.fit)
resfinder_M0 <- ggplot(dfA, aes(x = country, y = Mean)) + scale_color_manual(values=cols) +
geom_line() +
geom_jitter(data = dfA, aes(x = country, y = ARG_SUM, color = country), size = 12, alpha = 1, width = 0.3) +
geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.5, lwd = 0.75) + geom_point(size = 0.9) +
theme_linedraw() +
theme(axis.text.x = element_text(angle = 0, size = 40, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
axis.title.y = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
legend.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 40, family = "Times")) +
labs(y = "ARGs normalized to 16S rRNA", x = "") +
guides(color = FALSE, alpha = FALSE) +
labs(title = "Sum relative abundance of ARGs (ResFinder)")
resfinder_M0 + stat_pvalue_manual(pvalues, label = "p", y.position = 2.3, step.increase = 0.05, tip.length = 0.01, size = 9)
#ggsave(filename = "resfinder_sum_M0_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Modelling MGE abundance
## Data exploration
df<-data.frame(ARG_SUM=sample_sums(resfinder_PHY_stat),
intI1_SUM=sample_sums(MGE_PHY_int_stat),
MGE_SUM=sample_sums(MGE_PHY_stat),
hospital_section=as.factor(sample_data(resfinder_PHY_stat)$hospital_section),
SSU_counts=as.factor(sample_data(resfinder_PHY_stat)$SSU_counts),
hospital=as.factor(sample_data(resfinder_PHY_stat)$hospital),
category=as.factor(sample_data(resfinder_PHY_stat)$category),
city=as.factor(sample_data(resfinder_PHY_stat)$city),
country=as.factor(sample_data(resfinder_PHY_stat)$country),
continent=as.factor(sample_data(resfinder_PHY_stat)$continent),
R1_rpoB_counts=as.factor(sample_data(resfinder_PHY_stat)$R1_rpoB_counts),
no_of_beds=as.factor(sample_data(resfinder_PHY_stat)$no_of_beds),
daily_patient_visits=as.factor(sample_data(resfinder_PHY_stat)$daily_patient_visits),
long=as.factor(sample_data(resfinder_PHY_stat)$long),
lat=as.factor(sample_data(resfinder_PHY_stat)$lat),
A260_280=as.numeric(sample_data(resfinder_PHY_stat)$A260_280),
DNA_ng_µl=as.numeric(sample_data(resfinder_PHY_stat)$DNA_ng_µl),
M_Seqs_trimmed=as.numeric(sample_data(resfinder_PHY_stat)$M_Seqs_trimmed))
df$SSU_counts <- as.character(df$SSU_counts)
df$SSU_counts <- as.numeric(df$SSU_counts)
df$R1_rpoB_counts <- as.character(df$R1_rpoB_counts)
df$R1_rpoB_counts <- as.numeric(df$R1_rpoB_counts)
df$daily_patient_visits <- as.character(df$daily_patient_visits)
df$daily_patient_visits <- as.numeric(df$daily_patient_visits)## Warning: NAs introduced by coerciondf$no_of_beds <- as.character(df$no_of_beds)
df$no_of_beds <- as.numeric(df$no_of_beds)## Warning: NAs introduced by coerciondf$long <- as.numeric(gsub(",", ".", gsub("\\.", "", df$long)))
df$lat <- as.numeric(gsub(",", ".", gsub("\\.", "", df$lat)))
#str(df)
# Simple model
M0 <- glm(MGE_SUM ~ country,
data = df, family="Gamma"(link="log"))
summary(M0)##
## Call:
## glm(formula = MGE_SUM ~ country, family = Gamma(link = "log"),
## data = df)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.4294 -0.5305 -0.1611 0.3034 1.1036
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.94260 0.11918 7.909 4.66e-11 ***
## countryBurkina Faso 0.02026 0.15699 0.129 0.898
## countryFinland 0.11567 0.24205 0.478 0.634
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for Gamma family taken to be 0.3550713)
##
## Null deviance: 24.303 on 66 degrees of freedom
## Residual deviance: 24.219 on 64 degrees of freedom
## AIC: 239.3
##
## Number of Fisher Scoring iterations: 5glht.M0 <- glht(M0, mcp(country = "Tukey"))
summary(glht(glht.M0))## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored##
## Simultaneous Tests for General Linear Hypotheses
##
## Linear Hypotheses:
## Estimate Std. Error z value Pr(>|z|)
## Burkina Faso - Benin == 0 0.02026 0.15699 0.129 0.991
## Finland - Benin == 0 0.11567 0.24205 0.478 0.879
## Finland - Burkina Faso == 0 0.09542 0.23415 0.407 0.911
## (Adjusted p values reported -- single-step method)pvalues <- tibble::tribble(
~group1, ~group2, ~p,
"Benin", "Burkina Faso", 0.991,
"Benin", "Finland", 0.879,
"Burkina Faso", "Finland", 0.911
)
pvalues## # A tibble: 3 x 3
## group1 group2 p
## <chr> <chr> <dbl>
## 1 Benin Burkina Faso 0.991
## 2 Benin Finland 0.879
## 3 Burkina Faso Finland 0.911dfA <- cbind(df, Mean = predict(M0, newdata = df, type = "response"), SE = predict(M0,
newdata = df, type = "response", se.fit = T)$se.fit)
cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
MGE_M0 <- ggplot(dfA, aes(x = country, y = Mean)) + scale_color_manual(values=cols) +
geom_line() +
geom_jitter(data = dfA, aes(x = country, y = MGE_SUM, color = country), size = 12, alpha = 1, width = 0.3) +
geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.5, lwd = 0.75) + geom_point(size = 0.9) +
theme_linedraw() +
theme(axis.text.x = element_text(angle = 0, size = 40, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
axis.title.y = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
legend.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 40, family = "Times")) +
labs(y = "MGEs normalized to 16S rRNA", x = "") +
guides(color = FALSE, alpha = FALSE) +
labs(title = "Sum relative abundance of MGEs (MGE database)")
MGE_M0 + stat_pvalue_manual(pvalues, label = "p", y.position = 7, step.increase = 0.05, tip.length = 0.01, size = 9)
#ggsave(filename = "MGE_sum_M0_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Modelling intI1 abundance
## Data exploration
df<-data.frame(ARG_SUM=sample_sums(resfinder_PHY_stat),
intI1_SUM=sample_sums(MGE_PHY_int_stat),
MGE_SUM=sample_sums(MGE_PHY_stat),
hospital_section=as.factor(sample_data(resfinder_PHY_stat)$hospital_section),
SSU_counts=as.factor(sample_data(resfinder_PHY_stat)$SSU_counts),
hospital=as.factor(sample_data(resfinder_PHY_stat)$hospital),
category=as.factor(sample_data(resfinder_PHY_stat)$category),
city=as.factor(sample_data(resfinder_PHY_stat)$city),
country=as.factor(sample_data(resfinder_PHY_stat)$country),
continent=as.factor(sample_data(resfinder_PHY_stat)$continent),
R1_rpoB_counts=as.factor(sample_data(resfinder_PHY_stat)$R1_rpoB_counts),
no_of_beds=as.factor(sample_data(resfinder_PHY_stat)$no_of_beds),
daily_patient_visits=as.factor(sample_data(resfinder_PHY_stat)$daily_patient_visits),
long=as.factor(sample_data(resfinder_PHY_stat)$long),
lat=as.factor(sample_data(resfinder_PHY_stat)$lat),
A260_280=as.numeric(sample_data(resfinder_PHY_stat)$A260_280),
DNA_ng_µl=as.numeric(sample_data(resfinder_PHY_stat)$DNA_ng_µl),
M_Seqs_trimmed=as.numeric(sample_data(resfinder_PHY_stat)$M_Seqs_trimmed))
df$SSU_counts <- as.character(df$SSU_counts)
df$SSU_counts <- as.numeric(df$SSU_counts)
df$R1_rpoB_counts <- as.character(df$R1_rpoB_counts)
df$R1_rpoB_counts <- as.numeric(df$R1_rpoB_counts)
df$daily_patient_visits <- as.character(df$daily_patient_visits)
df$daily_patient_visits <- as.numeric(df$daily_patient_visits)## Warning: NAs introduced by coerciondf$no_of_beds <- as.character(df$no_of_beds)
df$no_of_beds <- as.numeric(df$no_of_beds)## Warning: NAs introduced by coerciondf$long <- as.numeric(gsub(",", ".", gsub("\\.", "", df$long)))
df$lat <- as.numeric(gsub(",", ".", gsub("\\.", "", df$lat)))
#str(df)
# Simple model
M0 <- glm(intI1_SUM ~ country,
data = df, family="Gamma"(link="log"))
summary(M0)##
## Call:
## glm(formula = intI1_SUM ~ country, family = Gamma(link = "log"),
## data = df)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.9566 -0.5599 -0.2306 0.3069 1.9490
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.0625 0.1479 -7.185 8.82e-10 ***
## countryBurkina Faso -0.5493 0.1948 -2.820 0.00639 **
## countryFinland -1.7648 0.3004 -5.876 1.65e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for Gamma family taken to be 0.5467518)
##
## Null deviance: 50.427 on 66 degrees of freedom
## Residual deviance: 35.229 on 64 degrees of freedom
## AIC: -82.305
##
## Number of Fisher Scoring iterations: 6glht.M0 <- glht(M0, mcp(country = "Tukey"))
summary(glht(glht.M0))## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored##
## Simultaneous Tests for General Linear Hypotheses
##
## Linear Hypotheses:
## Estimate Std. Error z value Pr(>|z|)
## Burkina Faso - Benin == 0 -0.5493 0.1948 -2.820 0.0126 *
## Finland - Benin == 0 -1.7648 0.3004 -5.876 <0.001 ***
## Finland - Burkina Faso == 0 -1.2155 0.2906 -4.183 <0.001 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)pvalues <- tibble::tribble(
~group1, ~group2, ~p,
"Benin", "Burkina Faso", 0.013,
"Benin", "Finland", 0.001,
"Burkina Faso", "Finland", 0.001
)
pvalues## # A tibble: 3 x 3
## group1 group2 p
## <chr> <chr> <dbl>
## 1 Benin Burkina Faso 0.013
## 2 Benin Finland 0.001
## 3 Burkina Faso Finland 0.001dfA <- cbind(df, Mean = predict(M0, newdata = df, type = "response"), SE = predict(M0,
newdata = df, type = "response", se.fit = T)$se.fit)
cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
MGE_M0 <- ggplot(dfA, aes(x = country, y = Mean)) + scale_color_manual(values=cols) +
geom_line() +
geom_jitter(data = dfA, aes(x = country, y = intI1_SUM, color = country), size = 12, alpha = 1, width = 0.3) +
geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.5, lwd = 0.75) + geom_point(size = 0.9) +
theme_linedraw() +
theme(axis.text.x = element_text(angle = 0, size = 40, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
axis.title.y = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
legend.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 40, family = "Times")) +
labs(y = "intI1 normalized to 16S rRNA", x = "") +
guides(color = FALSE, alpha = FALSE) +
labs(title = "Sum relative abundance of Class 1 integrons (MGE database)")
MGE_M0 + stat_pvalue_manual(pvalues, label = "p", y.position = 1.05, step.increase = 0.05, tip.length = 0.01, size = 9)
#ggsave(filename = "intI1_sum_M0_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Relative sums of ARGs with all samples based on category
# Subset samples which belong to category among >2 other samples
resfinder_PHY_cat_stat <- subset_samples(resfinder_PHY, category != "WA street gutter" &
category != "WA street gutter, sediment" &
category != "WA soil inbetween tanks" &
category != "WA empty hospital septic tank" &
category != "WA treated, receiving municipality channel" &
category != "WA treated, drinking water" &
category != "WA treated, receiving river" &
category != "WA treated, hand washing" &
category != "WA in-patient feces")
MGE_PHY_int_cat_stat <- subset_samples(MGE_PHY_int, category != "WA street gutter" &
category != "WA street gutter, sediment" &
category != "WA soil inbetween tanks" &
category != "WA empty hospital septic tank" &
category != "WA treated, receiving municipality channel" &
category != "WA treated, drinking water" &
category != "WA treated, receiving river" &
category != "WA treated, hand washing" &
category != "WA in-patient feces")
df<-data.frame(ARG_SUM=sample_sums(resfinder_PHY_cat_stat),
intI1_SUM=sample_sums(MGE_PHY_int_cat_stat),
hospital_section=as.factor(sample_data(resfinder_PHY_cat_stat)$hospital_section),
SSU_counts=as.factor(sample_data(resfinder_PHY_cat_stat)$SSU_counts),
hospital=as.factor(sample_data(resfinder_PHY_cat_stat)$hospital),
category=as.factor(sample_data(resfinder_PHY_cat_stat)$category),
city=as.factor(sample_data(resfinder_PHY_cat_stat)$city),
country=as.factor(sample_data(resfinder_PHY_cat_stat)$country),
continent=as.factor(sample_data(resfinder_PHY_cat_stat)$continent),
R1_rpoB_counts=as.factor(sample_data(resfinder_PHY_cat_stat)$R1_rpoB_counts),
no_of_beds=as.factor(sample_data(resfinder_PHY_cat_stat)$no_of_beds),
daily_patient_visits=as.factor(sample_data(resfinder_PHY_cat_stat)$daily_patient_visits),
sample_material=as.factor(sample_data(resfinder_PHY_cat_stat)$sample_material))
df$SSU_counts <- as.character(df$SSU_counts)
df$SSU_counts <- as.numeric(df$SSU_counts)
df$R1_rpoB_counts <- as.character(df$R1_rpoB_counts)
df$R1_rpoB_counts <- as.numeric(df$R1_rpoB_counts)
df$daily_patient_visits <- as.character(df$daily_patient_visits)
df$daily_patient_visits <- as.numeric(df$daily_patient_visits)## Warning: NAs introduced by coerciondf$no_of_beds <- as.character(df$no_of_beds)
df$no_of_beds <- as.numeric(df$no_of_beds)## Warning: NAs introduced by coercion#str(df)
# Simple model
M0 <- glm(ARG_SUM ~ category,
data = df, family="Gamma"(link="log"))
summary(M0)##
## Call:
## glm(formula = ARG_SUM ~ category, family = Gamma(link = "log"),
## data = df)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.47898 -0.43900 0.00759 0.18114 1.07417
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.04278 0.17261 0.248 0.80477
## categoryEu other wwtp influent -1.17222 0.25268 -4.639 1.11e-05 ***
## categoryNorth Eu hospital effluent -0.80439 0.24411 -3.295 0.00138 **
## categoryNorth Eu treated -3.11880 0.33053 -9.436 2.42e-15 ***
## categoryNorth Eu wwtp influent -1.66328 0.21374 -7.782 8.20e-12 ***
## categoryWA hospital effluent -0.04340 0.18394 -0.236 0.81397
## categoryWA river, drinking water -3.49161 0.33053 -10.564 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for Gamma family taken to be 0.2383578)
##
## Null deviance: 78.865 on 102 degrees of freedom
## Residual deviance: 24.955 on 96 degrees of freedom
## AIC: 26.328
##
## Number of Fisher Scoring iterations: 5# Predictions
glht.M0 <- glht(M0, mcp(category = "Tukey"))
summary(glht(glht.M0))## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored## Warning in chkdots(...): Argument(s) 'complete' passed to '...' are ignored##
## Simultaneous Tests for General Linear Hypotheses
##
## Linear Hypotheses:
## Estimate
## Eu other wwtp influent - Eu other hospital effluent == 0 -1.1722
## North Eu hospital effluent - Eu other hospital effluent == 0 -0.8044
## North Eu treated - Eu other hospital effluent == 0 -3.1188
## North Eu wwtp influent - Eu other hospital effluent == 0 -1.6633
## WA hospital effluent - Eu other hospital effluent == 0 -0.0434
## WA river, drinking water - Eu other hospital effluent == 0 -3.4916
## North Eu hospital effluent - Eu other wwtp influent == 0 0.3678
## North Eu treated - Eu other wwtp influent == 0 -1.9466
## North Eu wwtp influent - Eu other wwtp influent == 0 -0.4911
## WA hospital effluent - Eu other wwtp influent == 0 1.1288
## WA river, drinking water - Eu other wwtp influent == 0 -2.3194
## North Eu treated - North Eu hospital effluent == 0 -2.3144
## North Eu wwtp influent - North Eu hospital effluent == 0 -0.8589
## WA hospital effluent - North Eu hospital effluent == 0 0.7610
## WA river, drinking water - North Eu hospital effluent == 0 -2.6872
## North Eu wwtp influent - North Eu treated == 0 1.4555
## WA hospital effluent - North Eu treated == 0 3.0754
## WA river, drinking water - North Eu treated == 0 -0.3728
## WA hospital effluent - North Eu wwtp influent == 0 1.6199
## WA river, drinking water - North Eu wwtp influent == 0 -1.8283
## WA river, drinking water - WA hospital effluent == 0 -3.4482
## Std. Error z value
## Eu other wwtp influent - Eu other hospital effluent == 0 0.2527 -4.639
## North Eu hospital effluent - Eu other hospital effluent == 0 0.2441 -3.295
## North Eu treated - Eu other hospital effluent == 0 0.3305 -9.436
## North Eu wwtp influent - Eu other hospital effluent == 0 0.2137 -7.782
## WA hospital effluent - Eu other hospital effluent == 0 0.1839 -0.236
## WA river, drinking water - Eu other hospital effluent == 0 0.3305 -10.564
## North Eu hospital effluent - Eu other wwtp influent == 0 0.2527 1.456
## North Eu treated - Eu other wwtp influent == 0 0.3369 -5.778
## North Eu wwtp influent - Eu other wwtp influent == 0 0.2235 -2.197
## WA hospital effluent - Eu other wwtp influent == 0 0.1952 5.784
## WA river, drinking water - Eu other wwtp influent == 0 0.3369 -6.884
## North Eu treated - North Eu hospital effluent == 0 0.3305 -7.002
## North Eu wwtp influent - North Eu hospital effluent == 0 0.2137 -4.018
## WA hospital effluent - North Eu hospital effluent == 0 0.1839 4.137
## WA river, drinking water - North Eu hospital effluent == 0 0.3305 -8.130
## North Eu wwtp influent - North Eu treated == 0 0.3088 4.714
## WA hospital effluent - North Eu treated == 0 0.2889 10.643
## WA river, drinking water - North Eu treated == 0 0.3986 -0.935
## WA hospital effluent - North Eu wwtp influent == 0 0.1412 11.474
## WA river, drinking water - North Eu wwtp influent == 0 0.3088 -5.921
## WA river, drinking water - WA hospital effluent == 0 0.2889 -11.934
## Pr(>|z|)
## Eu other wwtp influent - Eu other hospital effluent == 0 <0.001 ***
## North Eu hospital effluent - Eu other hospital effluent == 0 0.0151 *
## North Eu treated - Eu other hospital effluent == 0 <0.001 ***
## North Eu wwtp influent - Eu other hospital effluent == 0 <0.001 ***
## WA hospital effluent - Eu other hospital effluent == 0 1.0000
## WA river, drinking water - Eu other hospital effluent == 0 <0.001 ***
## North Eu hospital effluent - Eu other wwtp influent == 0 0.7475
## North Eu treated - Eu other wwtp influent == 0 <0.001 ***
## North Eu wwtp influent - Eu other wwtp influent == 0 0.2718
## WA hospital effluent - Eu other wwtp influent == 0 <0.001 ***
## WA river, drinking water - Eu other wwtp influent == 0 <0.001 ***
## North Eu treated - North Eu hospital effluent == 0 <0.001 ***
## North Eu wwtp influent - North Eu hospital effluent == 0 0.0011 **
## WA hospital effluent - North Eu hospital effluent == 0 <0.001 ***
## WA river, drinking water - North Eu hospital effluent == 0 <0.001 ***
## North Eu wwtp influent - North Eu treated == 0 <0.001 ***
## WA hospital effluent - North Eu treated == 0 <0.001 ***
## WA river, drinking water - North Eu treated == 0 0.9616
## WA hospital effluent - North Eu wwtp influent == 0 <0.001 ***
## WA river, drinking water - North Eu wwtp influent == 0 <0.001 ***
## WA river, drinking water - WA hospital effluent == 0 <0.001 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)# Plot
dfA <- cbind(df, Mean = predict(M0, newdata = df, type = "response"), SE = predict(M0,
newdata = df, type = "response", se.fit = T)$se.fit)
cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 7)
cat_M0 <- ggplot(dfA, aes(x = category, y = Mean)) + scale_color_manual(values=cols) +
geom_line() +
geom_jitter(data = dfA, aes(x = category, y = ARG_SUM, color = category), size = 12, alpha = 1, width = 0.3) +
geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.5, lwd = 0.75) + geom_point(size = 0.9) +
theme_linedraw() +
theme(axis.text.x = element_text(angle = 45, size = 12, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 28, family = "Times"),
axis.title.y = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
legend.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 40, family = "Times")) +
labs(y = "IntI1 normalized to 16S rRNA", x = "") +
guides(color = FALSE, alpha = FALSE) +
labs(title = "Sum relative abundance of ARGs (resfinder)")
cat_M0
# Reorder samples
sam_order <- c("WA hospital effluent",
"Eu other hospital effluent",
"North Eu hospital effluent",
"Eu other wwtp influent",
"North Eu wwtp influent",
"WA river, drinking water",
"North Eu treated")
cat_M0$data$category <- factor(cat_M0$data$category, levels = sam_order)
cat_M0
#ggsave(filename = "resfinder_sum_categories_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Diversity
ResFinder
# Create phyloseq object with the count data
OTU_resfinder_counts <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
match <- match(rownames(metadata), colnames(OTU_resfinder_counts))
OTU_resfinder_counts <- OTU_resfinder_counts[,match]
all(colnames(OTU_resfinder_counts) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
match <- match(rownames(OTU_resfinder_counts), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder_counts) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Reload to get new row numbers
write.table(OTU_resfinder_counts, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_counts.txt",
row.names=T, sep = "\t", col.names = T)
OTU_resfinder_counts <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_resfinder_counts.txt", row.names=NULL)
OTU_resfinder_counts$row.names<-NULL
# Reload to get new row numbers
write.table(clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=NULL)
resfinder_PHY_counts <- phyloseq(otu_table(OTU_resfinder_counts, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
resfinder_PHY_counts = subset_samples(resfinder_PHY_counts, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
resfinder_PHY_counts <- subset_samples(resfinder_PHY_counts, country != "UK")
## Exclude biological / technical replicates
resfinder_PHY_counts_stat <- subset_samples(resfinder_PHY_counts, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
resfinder_PHY_counts_stat <- subset_samples(resfinder_PHY_counts_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
PHY <- prune_taxa(taxa_sums(resfinder_PHY_counts_stat) > 0, resfinder_PHY_counts_stat)
# Rarefy
# 90% of the minimum sample depth
rarecurve(t(otu_table(PHY)), step=50, cex=0.5)
PHY_rarefied = rarefy_even_depth(PHY, rngseed=1)## `set.seed(1)` was used to initialize repeatable random subsampling.## Please record this for your records so others can reproduce.## Try `set.seed(1); .Random.seed` for the full vector## ...## 567OTUs were removed because they are no longer
## present in any sample after random subsampling## ...#567OTUs were removed because they are no longer
#present in any sample after random subsampling
rarecurve(t(otu_table(PHY_rarefied)), step=50, cex=0.5)
tab <- microbiome::alpha(PHY_rarefied, index = "all")## Observed richness## Other forms of richness## Diversity## Evenness## Dominance## Rarity#kable(head(tab))
PHY.meta <- meta(PHY_rarefied)
#kable(head(PHY.meta))
PHY.meta$Shannon <- tab$diversity_shannon
PHY.meta$Observed <- tab$observed
# make a pairwise list that we want to compare.
vars <- levels(PHY.meta$country)
vars.pairs <- combn(seq_along(vars), 2, simplify = FALSE, FUN = function(i)vars[i])
print(vars.pairs)## [[1]]
## [1] "Benin" "Burkina Faso"
##
## [[2]]
## [1] "Benin" "Finland"
##
## [[3]]
## [1] "Burkina Faso" "Finland"cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
shannon <- ggboxplot(PHY.meta, x = "country", y = "Shannon", fill = "country", add = "jitter") +
scale_fill_manual(values = cols) +
theme(axis.text.x = element_text(angle = 0, vjust = 0.75, size = 36, hjust = 0.75, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
legend.position = "none",
legend.title = element_blank(),
axis.title.y = element_text(size = 36, family = "Times"),
plot.title = element_text(size = 40, family = "Times", face = "bold")) +
labs(title = "Alpha diversity, Shannon index (ResFinder)")
a <- shannon + stat_compare_means(comparisons = vars.pairs,
label = "p",
hide.ns = F, ref.group = ".all.",
size = 10, fontfamily = "Times")
a
#ggsave(filename = "resfinder_diversities_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Diversity
Metaphlan3
# transform into counts
PHY <- transform_sample_counts(metaphlan_PHY_stat, function(x) 1E6 * x/sum(x))
otu_table(PHY) <- round(otu_table(PHY), digits = 0)
PHY <- prune_taxa(taxa_sums(PHY) > 0, PHY)
# Sequences should be rarefied prior running Metaphlan3!
tab <- microbiome::alpha(PHY, index = "all")## Observed richness## Other forms of richness## Diversity## Evenness## Dominance## Rarity#kable(head(tab))
PHY.meta <- meta(PHY)
#kable(head(PHY.meta))
PHY.meta$Shannon <- tab$diversity_shannon
# make a pairwise list that we want to compare.
vars <- levels(PHY.meta$country)
vars.pairs <- combn(seq_along(vars), 2, simplify = FALSE, FUN = function(i)vars[i])
print(vars.pairs)## [[1]]
## [1] "Benin" "Burkina Faso"
##
## [[2]]
## [1] "Benin" "Finland"
##
## [[3]]
## [1] "Burkina Faso" "Finland"cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
shannon <- ggboxplot(PHY.meta, x = "country", y = "Shannon", fill = "country", add = "jitter") +
scale_fill_manual(values = cols) +
theme(axis.text.x = element_text(angle = 0, vjust = 0.75, size = 36, hjust = 0.75, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
legend.position = "none",
legend.title = element_blank(),
axis.title.y = element_text(size = 36, family = "Times"),
plot.title = element_text(size = 40, family = "Times", face = "bold")) +
labs(title = "Alpha diversity, Shannon index (Metaphlan3)")
a <- shannon + stat_compare_means(comparisons = vars.pairs,
label = "p",
hide.ns = F, ref.group = ".all.",
size = 10, fontfamily = "Times")
a
#ggsave(filename = "metaphlan_diversities_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Diversity
Mobilome (MGE database)
# Create phyloseq object with the count data
OTU_MGE_counts <-as.matrix(read.table("cp_MGE_genemat.txt", header= T, check.names = F, row.names = 1))
match <- match(rownames(metadata), colnames(OTU_MGE_counts))
OTU_MGE_counts <- OTU_MGE_counts[,match]
all(colnames(OTU_MGE_counts) == rownames(metadata))## [1] TRUE# Tax table
MGE_tax_table_trim <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", header=FALSE)
colnames(MGE_tax_table_trim) <- c("Gene", "Element", "Class")
# Reorder tax_table to match OTU_MGE
match <- match(rownames(OTU_MGE_counts), MGE_tax_table_trim$Gene)
MGE_tax_table_trim <- MGE_tax_table_trim[match,]
all(rownames(OTU_MGE_counts) == MGE_tax_table_trim$Gene)## [1] TRUE# Reload to get new row numbers
write.table(OTU_MGE_counts, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_MGE_counts",
row.names=T, sep = "\t", col.names = T)
OTU_MGE_counts <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/OTU_MGE_counts", row.names=NULL)
OTU_MGE_counts$row.names<-NULL
# Reload to get new row numbers
write.table(MGE_tax_table_trim, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", row.names=F, sep = "\t", col.names = T)
MGE_tax_table_trim <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/MGE_tax_table_trim.txt", row.names=NULL)
MGE_PHY_counts <- phyloseq(otu_table(OTU_MGE_counts, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(MGE_tax_table_trim)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
MGE_PHY_counts = subset_samples(MGE_PHY_counts, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
MGE_PHY_counts <- subset_samples(MGE_PHY_counts, country != "UK")
## Exclude biological / technical replicates
MGE_PHY_counts_stat <- subset_samples(MGE_PHY_counts, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
MGE_PHY_counts_stat <- subset_samples(MGE_PHY_counts_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
PHY <- prune_taxa(taxa_sums(MGE_PHY_counts_stat) > 0, MGE_PHY_counts_stat)
# Rarefy
# 90% of the minimum sample depth
rarecurve(t(otu_table(PHY)), step=50, cex=0.5)
PHY_rarefied = rarefy_even_depth(PHY, rngseed=1)## `set.seed(1)` was used to initialize repeatable random subsampling.## Please record this for your records so others can reproduce.## Try `set.seed(1); .Random.seed` for the full vector## ...## 489OTUs were removed because they are no longer
## present in any sample after random subsampling## ...# 489OTUs were removed because they are no longer
# present in any sample after random subsampling
tab <- microbiome::alpha(PHY_rarefied, index = "all")## Observed richness## Other forms of richness## Diversity## Evenness## Dominance## Rarity#kable(head(tab))
PHY.meta <- meta(PHY_rarefied)
#kable(head(PHY.meta))
PHY.meta$Shannon <- tab$diversity_shannon
# make a pairwise list that we want to compare.
vars <- levels(PHY.meta$country)
vars.pairs <- combn(seq_along(vars), 2, simplify = FALSE, FUN = function(i)vars[i])
print(vars.pairs)## [[1]]
## [1] "Benin" "Burkina Faso"
##
## [[2]]
## [1] "Benin" "Finland"
##
## [[3]]
## [1] "Burkina Faso" "Finland"cols <- get_palette(c("#B2182B", "#44AA99", "#2166AC"), 3)
#"#81A88D", "#B2182B", "#85D4E3", "#F4A582", "#4393C3", "#2166AC", "#D6604D"
shannon <- ggboxplot(PHY.meta, x = "country", y = "Shannon", fill = "country", add = "jitter") +
scale_fill_manual(values = cols) +
theme(axis.text.x = element_text(angle = 0, vjust = 0.75, size = 36, hjust = 0.75, family = "Times", face = "bold"),
axis.title.x = element_blank(),
axis.text.y = element_text(size = 36, family = "Times"),
legend.position = "none",
legend.title = element_blank(),
axis.title.y = element_text(size = 36, family = "Times"),
plot.title = element_text(size = 40, family = "Times", face = "bold")) +
labs(title = "Alpha diversity, Shannon index (MGE)")
a <- shannon + stat_compare_means(comparisons = vars.pairs,
label = "p",
hide.ns = F, ref.group = ".all.",
size = 10, fontfamily = "Times")
a
#ggsave(filename = "mge_diversities_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Ordination, ARGs (ResFinder)
resfinder_PHY_ord <- ordinate(resfinder_PHY_stat, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(resfinder_PHY_stat, resfinder_PHY_ord, color = "country")
resfinder.p_ord <- p_ord +
scale_color_manual(values = c("#B2182B", "#44AA99", "#2166AC")) +
geom_point(size = 5.5) +
stat_ellipse(level = 0.90, linetype = 1) +
geom_text_repel(mapping = aes(label = hospital), size = 6, family = "Times", hjust = 1.2) +
theme_minimal() + labs(title= "Resistome (ResFinder)", subtitle = "Hospital WWs in Benin, Burkina Faso and Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 20, family = "Times"),
legend.text = element_text(size = 36, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 36, family = "Times"),
axis.text = element_text(size = 18, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
resfinder.p_ord## Warning: ggrepel: 21 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
# Save
#ggsave(filename = "ord_resfinder_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Test significance using pair-wise adonis
#resfinder_temp <- subset_samples(resfinder_PHY_stat_equal, (country == "Benin" | country == "Finland"))
#resfinder_dist <- vegdist(t(otu_table(resfinder_temp)), dist = "horn")
#adonis(resfinder_dist ~ country, data = data.frame(sample_data(resfinder_temp), permutations = 9999))
#resfinder_temp <- subset_samples(resfinder_PHY_stat_equal, (country == "Benin" | country == "Burkina Faso"))
#resfinder_dist <- vegdist(t(otu_table(resfinder_temp)), dist = "horn")
#adonis(resfinder_dist ~ country, data = data.frame(sample_data(resfinder_temp), permutations = 9999))
#resfinder_temp <- subset_samples(resfinder_PHY_stat_equal, (country == "Burkina Faso" | country == "Finland"))
#resfinder_dist <- vegdist(t(otu_table(resfinder_temp)), dist = "horn")
#adonis(resfinder_dist ~ country, data = data.frame(sample_data(resfinder_temp), permutations = 9999))Ordination, taxa (Metaphlan3)
PHY = transform_sample_counts(metaphlan_PHY_stat, function(x) 1E6 * x/sum(x))
metaphlan_PHY_ord <- ordinate(PHY, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(PHY, metaphlan_PHY_ord, color = "country")
metaphlan.p_ord <- p_ord +
scale_color_manual(values = c("#B2182B", "#44AA99", "#2166AC")) +
geom_point(size = 5.5) +
stat_ellipse(level = 0.90, linetype = 1) +
geom_text_repel(mapping = aes(label = hospital), size = 6, family = "Times", hjust = 1.2) +
theme_minimal() + labs(title= "Taxonomy (Metaphlan3)", subtitle = "Hospital WWs in Benin, Burkina Faso and Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 20, family = "Times"),
legend.text = element_text(size = 50, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 36, family = "Times"),
axis.text = element_text(size = 18, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
metaphlan.p_ord## Warning: ggrepel: 38 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
# Save
#ggsave(filename = "ord_metaphlan_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Test significance using pair-wise adonis
#metaphlan_temp <- subset_samples(metaphlan_PHY_stat_equal, (country == "Benin" | country == "Finland"))
#metaphlan_dist <- vegdist(t(otu_table(metaphlan_temp)), dist = "horn")
#adonis(metaphlan_dist ~ country, data = data.frame(sample_data(metaphlan_temp), permutations = 9999))
#metaphlan_temp <- subset_samples(metaphlan_PHY_stat_equal, (country == "Benin" | country == "Burkina Faso"))
#metaphlan_dist <- vegdist(t(otu_table(metaphlan_temp)), dist = "horn")
#adonis(metaphlan_dist ~ country, data = data.frame(sample_data(metaphlan_temp), permutations = 9999))
#metaphlan_temp <- subset_samples(metaphlan_PHY_stat_equal, (country == "Burkina Faso" | country == "Finland"))
#metaphlan_dist <- vegdist(t(otu_table(metaphlan_temp)), dist = "horn")
#adonis(metaphlan_dist ~ country, data = data.frame(sample_data(metaphlan_temp), permutations = 9999))Ordination, MGEs (Mobilome)
MGE_PHY_ord <- ordinate(MGE_PHY_stat, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(MGE_PHY_stat, MGE_PHY_ord, color = "country")
MGE.p_ord <- p_ord +
scale_color_manual(values = c("#B2182B", "#44AA99", "#2166AC")) +
geom_point(size = 5.5) +
stat_ellipse(level = 0.90, linetype = 1) +
geom_text_repel(mapping = aes(label = hospital), size = 6, family = "Times", hjust = 1.2) +
theme_minimal() + labs(title= "Mobilome (MGE database)", subtitle = "Hospital WWs in Benin, Burkina Faso and Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 20, family = "Times"),
legend.text = element_text(size = 50, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 36, family = "Times"),
axis.text = element_text(size = 18, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
MGE.p_ord## Warning: ggrepel: 43 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
# Save
#ggsave(filename = "ord_mge_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Test significance using pair-wise adonis
#MGE_temp <- subset_samples(MGE_PHY_stat_equal, (country == "Benin" | country == "Finland"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))
#MGE_temp <- subset_samples(MGE_PHY_stat_equal, (country == "Benin" | country == "Burkina Faso"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))
#MGE_temp <- subset_samples(MGE_PHY_stat_equal, (country == "Burkina Faso" | country == "Finland"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))Ordination, intI1 (Mobilome)
intI1_PHY_ord <- ordinate(MGE_PHY_int_stat, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(MGE_PHY_int_stat, intI1_PHY_ord, color = "country")
intI1.p_ord <- p_ord +
scale_color_manual(values = c("#B2182B", "#44AA99", "#2166AC")) +
geom_point(size = 5.5) +
stat_ellipse(level = 0.90, linetype = 1) +
geom_text_repel(mapping = aes(label = hospital), size = 3, family = "Times", hjust = 1.2) +
theme_minimal() + labs(title= "Mobilome, IntI1 (MGE Database)", subtitle = "Hospital WWs in Benin, Burkina Faso and Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
intI1.p_ord## Warning: ggrepel: 42 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
# Save
#ggsave(filename = "ord_intI1_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Test significance using pair-wise adonis
#MGE_temp <- subset_samples(MGE_PHY_int_stat_equal, (country == "Benin" | country == "Finland"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))
#MGE_temp <- subset_samples(MGE_PHY_int_stat_equal, (country == "Benin" | country == "Burkina Faso"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))
#MGE_temp <- subset_samples(MGE_PHY_int_stat_equal, (country == "Burkina Faso" | country == "Finland"))
#MGE_dist <- vegdist(t(otu_table(MGE_temp)), dist = "horn")
#adonis(MGE_dist ~ country, data = data.frame(sample_data(MGE_temp), permutations = 9999))Ordination, VFDBs (Virulence genes)
VFDB_PHY_ord <- ordinate(VFDB_PHY_stat, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(VFDB_PHY_stat, VFDB_PHY_ord, color = "country")
VFDB.p_ord <- p_ord +
scale_color_manual(values = c("#B2182B", "#44AA99", "#2166AC")) +
geom_point(size = 5.5) +
stat_ellipse(level = 0.90, linetype = 1) +
# geom_text_repel(mapping = aes(label = hospital), size = 3, family = "Times", hjust = 1.2) +
theme_minimal() + labs(title= "Virulence Genes (VFDB)", subtitle = "Hospital WWs in Benin (25), BF (34) and Finland (8)") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
VFDB.p_ord
# Save
#ggsave(filename = "ord_vfdb_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Test significance using pair-wise adonis
#VFDB_temp <- subset_samples(VFDB_PHY_stat_equal, (country == "Benin" | country == "Finland"))
#VFDB_dist <- vegdist(t(otu_table(VFDB_temp)), dist = "horn")
#adonis(VFDB_dist ~ country, data = data.frame(sample_data(VFDB_temp), permutations = 9999))
#VFDB_temp <- subset_samples(VFDB_PHY_stat_equal, (country == "Benin" | country == "Burkina Faso"))
#VFDB_dist <- vegdist(t(otu_table(VFDB_temp)), dist = "horn")
#adonis(VFDB_dist ~ country, data = data.frame(sample_data(VFDB_temp), permutations = 9999))
#VFDB_temp <- subset_samples(VFDB_PHY_stat_equal, (country == "Burkina Faso" | country == "Finland"))
#VFDB_dist <- vegdist(t(otu_table(VFDB_temp)), dist = "horn")
#adonis(VFDB_dist ~ country, data = data.frame(sample_data(VFDB_temp), permutations = 9999))Compare hospitals within each country
Ordination, ARGs (ResFinder)
resfinder_PHY_WA_WW_stat_Ben <- subset_samples(resfinder_PHY_WA_WW_stat, country == "Benin")
cols <- get_palette(c("#332288", "#44AA99", "#88CCEE", "#B94ED9", "#F22D3D"), 5)
resfinder_PHY_ord <- ordinate(resfinder_PHY_WA_WW_stat_Ben, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(resfinder_PHY_WA_WW_stat_Ben, resfinder_PHY_ord, color = "hospital")
resfinder.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Resistome (ResFinder)", subtitle = "Hospital WWs in Benin") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
resfinder.p_ord## Too few points to calculate an ellipse## Warning: Removed 1 row(s) containing missing values (geom_path).
# Save
#ggsave(filename = "ord_res_Ben_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
resfinder_PHY_WA_WW_stat_BF <- subset_samples(resfinder_PHY_WA_WW_stat, country == "Burkina Faso")
resfinder_PHY_ord <- ordinate(resfinder_PHY_WA_WW_stat_BF, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(resfinder_PHY_WA_WW_stat_BF, resfinder_PHY_ord, color = "hospital")
resfinder.p_ord <- p_ord +
scale_color_manual(values=cols) +
scale_shape_manual(values=c(18, 16)) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Resistome (ResFinder)", subtitle = "Hospital WWs in Burkina Faso") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
resfinder.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 2 row(s) containing missing values (geom_path).
# Save
#ggsave(filename = "ord_res_BF_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
resfinder_PHY_stat_Fin <- subset_samples(resfinder_PHY_stat, category == "North Eu hospital effluent")
resfinder_PHY_ord <- ordinate(resfinder_PHY_stat_Fin, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(resfinder_PHY_stat_Fin, resfinder_PHY_ord, color = "hospital")
resfinder.p_ord <- p_ord +
scale_color_manual(values=cols) +
scale_shape_manual(values=c(18, 16)) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Resistome (ResFinder)", subtitle = "Hospital WWs in Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
resfinder.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 4 row(s) containing missing values (geom_path).
# Save
#ggsave(filename = "ord_res_Fin_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)WA WW samples
Ordination, taxa (Metaphlan3)
metaphlan_PHY_WA_WW_stat_Ben <- subset_samples(metaphlan_PHY_WA_WW_stat, country == "Benin")
metaphlan_PHY_ord <- ordinate(metaphlan_PHY_WA_WW_stat_Ben, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(metaphlan_PHY_WA_WW_stat_Ben, metaphlan_PHY_ord, color = "hospital")
metaphlan.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Taxonomy (Metaphlan3)", subtitle = "Hospital WWs in Benin") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
metaphlan.p_ord## Too few points to calculate an ellipse## Warning: Removed 1 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_Ben_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
metaphlan_PHY_WA_WW_stat_BF <- subset_samples(metaphlan_PHY_WA_WW_stat, country == "Burkina Faso")
metaphlan_PHY_ord <- ordinate(metaphlan_PHY_WA_WW_stat_BF, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(metaphlan_PHY_WA_WW_stat_BF, metaphlan_PHY_ord, color = "hospital")
metaphlan.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Taxonomy (Metaphlan3)", subtitle = "Hospital WWs in BF") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
metaphlan.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 2 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_BF_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
metaphlan_PHY_WA_WW_stat_Fin <- subset_samples(metaphlan_PHY, category == "North Eu hospital effluent")
metaphlan_PHY_ord <- ordinate(metaphlan_PHY_WA_WW_stat_Fin, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(metaphlan_PHY_WA_WW_stat_Fin, metaphlan_PHY_ord, color = "hospital")
metaphlan.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Taxonomy (Metaphlan3)", subtitle = "Hospital WWs in Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
metaphlan.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 4 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_Fin_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)WA WW samples
Ordination, mobilome (MGE database)
MGE_PHY_WA_WW_stat_Ben <- subset_samples(MGE_PHY_WA_WW_stat, country == "Benin")
MGE_PHY_ord <- ordinate(MGE_PHY_WA_WW_stat_Ben, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(MGE_PHY_WA_WW_stat_Ben, MGE_PHY_ord, color = "hospital")
MGE.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
geom_text(mapping = aes(label = hospital), size = 3, hjust = 1.5, vjust = -0.95) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Mobilome (MGE database)", subtitle = "Hospital WWs in Benin") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
MGE.p_ord## Too few points to calculate an ellipse## Warning: Removed 1 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_mge_Ben_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
MGE_PHY_WA_WW_stat_BF <- subset_samples(MGE_PHY_WA_WW_stat, country == "Burkina Faso")
MGE_PHY_ord <- ordinate(MGE_PHY_WA_WW_stat_BF, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(MGE_PHY_WA_WW_stat_BF, MGE_PHY_ord, color = "hospital")
MGE.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
geom_text(mapping = aes(label = hospital), size = 3, hjust = 1.5, vjust = -0.95) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Mobilome (MGE database)", subtitle = "Hospital WWs in BF") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
MGE.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 2 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_mge_BF_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
MGE_PHY_WA_WW_stat_Fin <- subset_samples(MGE_PHY, category == "North Eu hospital effluent")
MGE_PHY_ord <- ordinate(MGE_PHY_WA_WW_stat_Fin, method = "PCoA", distance = "horn")
p_ord <- plot_ordination(MGE_PHY_WA_WW_stat_Fin, MGE_PHY_ord, color = "hospital")
MGE.p_ord <- p_ord +
scale_color_manual(values=cols) +
geom_point(size = 7) +
geom_text(mapping = aes(label = hospital), size = 3, hjust = 1.5, vjust = -0.95) +
stat_ellipse(level = 0.90, linetype = 1) +
theme_minimal() + labs(title= "Mobilome (MGE database)", subtitle = "Hospital WWs in Finland") +
theme(plot.title = element_text(size = 36, family = "Times", face = "bold"),
plot.subtitle = element_text(size = 28, family = "Times"),
legend.text = element_text(size = 28, family = "Times"),
legend.title = element_blank(),
axis.title = element_text(size = 28, family = "Times")) +
guides(fill = guide_legend(override.aes = list(linetype = 0)),
color = guide_legend(override.aes = list(linetype = 0, size=5)))
MGE.p_ord## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse
## Too few points to calculate an ellipse## Warning: Removed 4 row(s) containing missing values (geom_path).
#ggsave(filename = "ord_mge_Fin_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)CrAssphage
# Plot orrelation of crAssphage & ARGs
ARG_relative_sum <- data.frame(sample_sums(resfinder_PHY_stat))
crAssphage_relative_sum <- data.frame(sample_sums(crass_PHY_stat))
all(rownames(ARG_relative_sum) == rownames(crAssphage_relative_sum))## [1] TRUE# Join data
crass_res <- cbind(ARG_relative_sum, crAssphage_relative_sum)
colnames(crass_res) <- c("ARGs", "crAssphage")
# Plot
cor <- ggplot(crass_res, aes(x=ARGs, y=crAssphage)) +
geom_point(size=7, shape=19, color = "#3110D2") +
geom_smooth(method="lm", se=TRUE, fullrange=FALSE, level=0.95, color = "#FB2A38", fill = "#8A91F8") +
theme_bw() +
theme(axis.title = element_text(size = 30, family = "Times"),
axis.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 30, family = "Times")) +
xlab("ARGs") + ylab("crAssphage") +
ggtitle("Correlation of relative sums of ARGs and crAssphage")
cor2 <- cor + stat_cor(method = "pearson", label.x = 0.1, label.y = 0)
cor2## `geom_smooth()` using formula 'y ~ x'
## Correlation between SSU & rpoB counts
## Use either *_PHY / *_PHY_stat_equal / *_PHY_WA_WW_stat
SSU_counts <- data.frame(sample_data(resfinder_PHY_stat)$SSU_counts)
R1_rpoB_counts <- data.frame(sample_data(resfinder_PHY_stat)$R1_rpoB_counts)
bacterial_counts <- cbind(SSU_counts, R1_rpoB_counts)
colnames(bacterial_counts) <- c("SSU_counts", "R1_rpoB_counts")
p <- ggplot(bacterial_counts, aes(x=SSU_counts, y=R1_rpoB_counts)) +
geom_point(size=7, shape=19, color = "#3110D2") +
geom_smooth(method="lm", se=TRUE, fullrange=FALSE, level=0.95, color = "#FB2A38", fill = "#8A91F8") +
theme_bw() +
theme(axis.title = element_text(size = 30, family = "Times"),
axis.text = element_text(size = 32, family = "Times"),
plot.title = element_text(size = 36, family = "Times"),
plot.subtitle = element_text(size = 28, family = "Times")) +
xlab("16s rRNA counts") + ylab("R1 rpoB counts") +
labs(title= "Correlation of 16s rRNA and rpoB counts", subtitle = "Hospital WWs in Benin (25), BF (34) and Finland (8)")
cor <- p + stat_cor(method = "pearson", label.x = 100000, label.y = 1000, )
cor## `geom_smooth()` using formula 'y ~ x'
correl<-corr.test(SSU_counts, R1_rpoB_counts, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(correl$r)
p <- data.frame(correl$p)
p.ad <- data.frame(correl$p.adj)
#ggsave(filename = "SSU_rpoB_cor_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Correlation of Intl1 & all ARGs
ARG_relative_sum <- data.frame(sample_sums(resfinder_PHY_stat))
MGE_relative_sum <- data.frame(sample_sums(MGE_PHY_stat))
all(rownames(ARG_relative_sum) == rownames(MGE_relative_sum))## [1] TRUE# Join data
mge_res <- cbind(ARG_relative_sum, MGE_relative_sum)
colnames(mge_res) <- c("ARGs", "MGEs")
# Plot
cor <- ggplot(mge_res, aes(x=ARGs, y=MGEs)) +
geom_point(size=7, shape=19, color = "#3110D2") +
geom_smooth(method="lm", se=TRUE, fullrange=FALSE, level=0.95, color = "#FB2A38", fill = "#8A91F8") +
theme_bw() +
theme(axis.title = element_text(size = 30, family = "Times"),
axis.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 36, family = "Times"),
plot.subtitle = element_text(size = 28, family = "Times")) +
xlab("ARG") + ylab("MGEs") +
labs(title= "Correlation of relative sums of ARGs and MGEs", subtitle = "Hospital WWs in Benin (25), BF (34) and Finland (8)")
cor2 <- cor + stat_cor(method = "pearson", label.x = 2, label.y = 1.5)
cor2## `geom_smooth()` using formula 'y ~ x'
#ggsave(filename = "ARG_MGE_cor_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
## Intl1
ARG_relative_sum <- data.frame(sample_sums(resfinder_PHY_stat))
intI1_relative_sum <- data.frame(sample_sums(MGE_PHY_int_stat))
all(rownames(ARG_relative_sum) == rownames(intI1_relative_sum))## [1] TRUE# Join data
intl_res <- cbind(ARG_relative_sum, intI1_relative_sum)
colnames(intl_res) <- c("ARGs", "intI1")
# Plot
cor <- ggplot(intl_res, aes(x=ARGs, y=intI1)) +
geom_point(size=7, shape=19, color = "#3110D2") +
geom_smooth(method="lm", se=TRUE, fullrange=FALSE, level=0.95, color = "#FB2A38", fill = "#8A91F8") +
theme_bw() +
theme(axis.title = element_text(size = 30, family = "Times"),
axis.text = element_text(size = 28, family = "Times"),
plot.title = element_text(size = 36, family = "Times"),
plot.subtitle = element_text(size = 28, family = "Times")) +
xlab("ARG") + ylab("intI1") +
labs(title= "Correlation of relative sums of ARGs and Int1", subtitle = "Hospital WWs in Benin, Burkina Faso and Finland")
cor2 <- cor + stat_cor(method = "pearson", label.x = 1, label.y = 1.5)
cor2## `geom_smooth()` using formula 'y ~ x'
correl<-corr.test(ARG_relative_sum, intI1_relative_sum, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(correl$r)
p <- data.frame(correl$p)
p.ad <- data.frame(correl$p.adj)
# Correlation coefficients for each country:
# Benin = 0.79, p = 2.5 x 10-6; BF=R2=0.54, p = 0.00086; Finland = 0.83, p = 0.01
#ggsave(filename = "ARG_intl1_cor_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)DESeq2
ResFinder
Fin-Ben
OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_resfinder))
OTU_resfinder <- OTU_resfinder[,match]
all(colnames(OTU_resfinder) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Divide by ARG gene lengths
## Get the lengths in terminal
# seqkit fx2tab --length --name --header-line resfinder.fasta > resfinder_lengths.txt
resfinder_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/resfinder_lengths.txt", header=FALSE, comment.char="#")
all(rownames(clusters_tax_table_resfinder$Gene) == resfinder_lengths$V1)## [1] TRUEOTU_resfinder_length_norm <- OTU_resfinder/resfinder_lengths[, 2]
# Normalization with Metaxa2 SSU counts
deseq_OTU_resfinder <- t(t(OTU_resfinder_length_norm)/metadata$SSU_counts) * 1540
all(rownames(metadata) == colnames(deseq_OTU_resfinder))## [1] TRUEidentical(OTU_resfinder_length_norm[2025, 5]/metadata$SSU_counts[5], deseq_OTU_resfinder[2025, 5])## [1] TRUEall(rownames(OTU_resfinder_length_norm) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Deseq
deseq_OTU <- deseq_OTU_resfinder[, ] * 10^5 + 1
# Reload to get new row numbers
write.table(deseq_OTU, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/deseq_OTU.txt",
row.names=T, sep = "\t", col.names = T)
deseq_OTU <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/deseq_OTU.txt", row.names=NULL)
deseq_OTU$row.names<-NULL
# Reload to get new row numbers
write.table(clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=NULL)
resfinder_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
resfinder_deseq = subset_samples(resfinder_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
resfinder_deseq <- subset_samples(resfinder_deseq, country != "UK")
## Exclude biological / technical replicates
resfinder_deseq <- subset_samples(resfinder_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
resfinder_deseq_stat <- subset_samples(resfinder_deseq, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Take pair wise comparisons
deseq_PHY = subset_samples(resfinder_deseq_stat, country == "Benin" | country == "Finland")
hist(log10(apply(otu_table(deseq_PHY), 1, var)), xlab = "log10(variance)")
# Let's set a threashold for the variance
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 936 taxa and 33 samples ]
## sample_data() Sample Data: [ 33 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 936 taxa by 3 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer modedds$category <- relevel(dds$country, "Benin", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## -- replacing outliers and refitting for 256 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## fitting model and testingres = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Benin"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_resfinder = res[which(res$padj < alpha), ]
sigtab_resfinder = cbind(as(sigtab_resfinder, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_resfinder),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_resfinder = merge(sigtab_resfinder, as.data.frame(n), by = 0)
#kable(sigtab_resfinder, caption = "Taxonomy")
# Plot
# Class
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Class = factor(as.character(sigtab_resfinder$Class), levels = names(x))
# Gene cluster
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Cluster_name, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Cluster_name = factor(as.character(sigtab_resfinder$Cluster_name), levels = names(x))
# Gene
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Gene, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Gene = factor(as.character(sigtab_resfinder$Gene), levels = names(x))
sorted_sigtab <- sigtab_resfinder[order(-sigtab_resfinder$log2FoldChange), ]
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/DESeq2_Ben_Fin.txt",
# row.names=T, sep = "\t", col.names = T)
Fin_Ben <- subset(sorted_sigtab,log2FoldChange>=0)
Ben_Fin <- subset(sorted_sigtab,log2FoldChange<=0)
Ben_Fin <- Ben_Fin[order(Ben_Fin$log2FoldChange), ]
head(Fin_Ben)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 6 sp1041 112.95961 8.097550 0.6553558 12.355960 4.523266e-35
## 465 sp480 51.72887 7.662155 0.6653165 11.516557 1.088754e-30
## 7 sp1042 105.26333 7.399562 0.6909062 10.709937 9.142365e-27
## 16 sp1096 73.09887 7.361624 0.7095942 10.374414 3.241945e-25
## 25 sp130 59.75044 7.296998 0.7752919 9.411935 4.870848e-21
## 362 sp285 42.59734 7.230515 0.5887217 12.281721 1.135570e-34
## padj Class Cluster_name Gene n
## 6 4.233777e-32 Betalactam blaOXA-280_1_clust blaOXA-211_1_JN861779 10
## 465 2.547685e-28 Betalactam blaOXA-299_1 blaOXA-299_1_APQD01000016 9
## 7 1.222465e-24 Betalactam blaOXA-280_1_clust blaOXA-212_1_JN861780 14
## 16 3.793076e-23 Betalactam blaOXA-280_1_clust blaOXA-334_1_KF203108 11
## 25 3.507011e-19 Aminoglycoside aac(6')-Ig_1 aac(6')-Ig_1_L09246 7
## 362 5.314465e-32 Betalactam blaOXA-280_1_clust blaOXA-373_1_HG931732 9head(Ben_Fin)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 253 sp2513 364.1142 -6.873014 0.6895513 -9.967372 2.117583e-23
## 300 sp2758 574.6104 -6.833083 0.6851255 -9.973476 1.991353e-23
## 301 sp2759 277.2942 -6.654446 0.7500006 -8.872588 7.146534e-19
## 78 sp154 577.0557 -6.015922 0.7130589 -8.436781 3.261972e-17
## 192 sp2207 163.2347 -6.001513 0.7831714 -7.663091 1.815109e-14
## 498 sp55 800.5257 -5.827984 0.6885997 -8.463529 2.594059e-17
## padj Class Cluster_name Gene n
## 253 1.651715e-21 Lincosamide lnu(F)_1_clust lnu(F)_3_AJ561197 27
## 300 1.651715e-21 Quinolone qnrVC4_1_clust qnrVC4_1_GQ891757 29
## 301 3.344578e-17 Quinolone qnrVC4_1_clust qnrVC5_1_JN408080 27
## 78 9.252139e-16 Aminoglycoside aac(6')-IIc_1 aac(6')-IIc_1_NC_012555 27
## 192 3.467228e-13 Betalactam blaCARB-50_1_clust blaCARB-2_1_M69058 26
## 498 8.093465e-16 Aminoglycoside ant(2'')-Ia_2_clust ant(2'')-Ia_6_AJ871915 29Plot
Benin-Finland
# Take top 25 from each and combine back to one table
#Fin25 <- Fin_Ben[1:25,]
#Ben25 <- Ben_Fin[1:25,]
#Ben_Fin_top_sigtab <- rbind(Fin25, Ben25)
# Shorten gene names and order again
#Ben_Fin_top_sigtab$Gene <- gsub(pattern = "_[A-Z].*", replacement = "", Ben_Fin_top_sigtab$Gene)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 13)
#p <- ggplot(data = Ben_Fin_top_sigtab,aes(x = Gene, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) +
# ylab("Log2 Fold Change\nBenin Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Enriched ARGs in Beninise / Finnish samples (25 with highest log 2 fold change values of each country)") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 10),
# axis.text.y = element_text(family = "Times", size = 12, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold"),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 20),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 20, face = "bold"))
#ggsave(filename = "BEN_FIN_deseq50_resfinder_genes_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1) DESeq2
ResFinder
BF-Fin
OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_resfinder))
OTU_resfinder <- OTU_resfinder[,match]
all(colnames(OTU_resfinder) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Divide by ARG gene lengths
## Get the lengths in terminal
# seqkit fx2tab --length --name --header-line resfinder.fasta > resfinder_lengths.txt
resfinder_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/resfinder_lengths.txt", header=FALSE, comment.char="#")
all(rownames(clusters_tax_table_resfinder$Gene) == resfinder_lengths$V1)## [1] TRUEOTU_resfinder_length_norm <- OTU_resfinder/resfinder_lengths[, 2]
# Normalization with Metaxa2 SSU counts
deseq_OTU_resfinder <- t(t(OTU_resfinder_length_norm)/metadata$SSU_counts) * 1540
all(rownames(metadata) == colnames(deseq_OTU_resfinder))## [1] TRUEidentical(OTU_resfinder_length_norm[2025, 5]/metadata$SSU_counts[5], deseq_OTU_resfinder[2025, 5])## [1] TRUEall(rownames(OTU_resfinder_length_norm) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Deseq
deseq_OTU <- deseq_OTU_resfinder[, ] * 10^5 + 1
# Reload to get new row numbers
write.table(deseq_OTU, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/deseq_OTU.txt",
row.names=T, sep = "\t", col.names = T)
deseq_OTU <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/deseq_OTU.txt", row.names=NULL)
deseq_OTU$row.names<-NULL
# Reload to get new row numbers
write.table(clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt", row.names=NULL)
resfinder_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
resfinder_deseq = subset_samples(resfinder_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
resfinder_deseq <- subset_samples(resfinder_deseq, country != "UK")
## Exclude biological / technical replicates
resfinder_deseq <- subset_samples(resfinder_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
resfinder_deseq_stat <- subset_samples(resfinder_deseq, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Take pair wise comparisons
deseq_PHY = subset_samples(resfinder_deseq_stat, country == "Burkina Faso" | country == "Finland")
hist(log10(apply(otu_table(deseq_PHY), 1, var)), xlab = "log10(variance)")
# Let's set a threashold for the variance
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 1099 taxa and 42 samples ]
## sample_data() Sample Data: [ 42 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 1099 taxa by 3 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer mode## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]dds$category <- relevel(dds$country, "Burkina Faso", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors
## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## final dispersion estimates## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## -- replacing outliers and refitting for 249 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]res = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Burkina.Faso"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_resfinder = res[which(res$padj < alpha), ]
sigtab_resfinder = cbind(as(sigtab_resfinder, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_resfinder),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_resfinder = merge(sigtab_resfinder, as.data.frame(n), by = 0)
#kable(sigtab_resfinder, caption = "Taxonomy")
# Plot
# Class
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Class = factor(as.character(sigtab_resfinder$Class), levels = names(x))
# Gene cluster
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Cluster_name, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Cluster_name = factor(as.character(sigtab_resfinder$Cluster_name), levels = names(x))
# Gene
x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Gene, function(x) max(x))
x = sort(x, TRUE)
sigtab_resfinder$Gene = factor(as.character(sigtab_resfinder$Gene), levels = names(x))
sorted_sigtab <- sigtab_resfinder[order(-sigtab_resfinder$log2FoldChange), ]
# Save BF
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/DESeq2_BF_Fin.txt",
# row.names=T, sep = "\t", col.names = T)
Fin_BF <- subset(sorted_sigtab,log2FoldChange>=0)
BF_Fin <- subset(sorted_sigtab,log2FoldChange<=0)
BF_Fin <- BF_Fin[order(BF_Fin$log2FoldChange), ]
head(Fin_BF)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 405 sp480 40.99270 7.852266 0.6026931 13.028630 8.410383e-39
## 15 sp1096 56.79119 7.700764 0.6040253 12.749075 3.154645e-37
## 92 sp1648 46.56010 7.387248 0.7266569 10.166075 2.810054e-24
## 14 sp1095 31.29971 7.021763 0.5779840 12.148714 5.827481e-34
## 8 sp1041 89.93935 6.883085 0.7940390 8.668447 4.380527e-18
## 19 sp130 46.71833 6.856940 0.7356964 9.320340 1.159679e-20
## padj Class Cluster_name Gene n
## 405 9.243010e-36 Betalactam blaOXA-299_1 blaOXA-299_1_APQD01000016 8
## 15 1.733478e-34 Betalactam blaOXA-280_1_clust blaOXA-334_1_KF203108 13
## 92 4.411785e-22 Betalactam blaOXA-296_1 blaOXA-296_1_APOH01000009 10
## 14 1.280880e-31 Betalactam blaOXA-280_1_clust blaOXA-333_1_KF203107 10
## 8 3.008875e-16 Betalactam blaOXA-280_1_clust blaOXA-211_1_JN861779 10
## 19 1.062073e-18 Aminoglycoside aac(6')-Ig_1 aac(6')-Ig_1_L09246 9head(BF_Fin)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 379 sp3033 343.1398 -6.688023 0.8193830 -8.162267 3.287957e-16
## 176 sp2103 607.6167 -6.365877 0.6600816 -9.644076 5.207786e-22
## 324 sp2850 328.3398 -5.761664 0.9097878 -6.332976 2.404772e-10
## 387 sp3068 180.0523 -5.539669 0.6107521 -9.070242 1.187535e-19
## 490 sp895 288.7270 -5.504616 0.8099697 -6.796076 1.075069e-11
## 496 sp948 801.0270 -5.411751 0.7492859 -7.222544 5.102375e-13
## padj Class Cluster_name Gene n
## 379 1.901824e-14 Trimethoprim dfrB1_1_clust dfrB5_1_AY943084 32
## 176 6.359285e-20 Betalactam blaCMY-150_2_clust blaCMY-4_1_LNHZ01000079 41
## 324 6.607111e-09 Sulphonamide sul4_1 sul4_1_MG649393 38
## 387 9.322147e-18 Trimethoprim dfrA15_4_clust dfrA15_2_AF221900 37
## 490 3.692191e-10 Betalactam blaOXA-46_1_clust blaOXA-46_1_AF317511 38
## 496 2.076856e-11 Betalactam blaOXA-368_1_clust blaOXA-101_1_AM412777 39Plot
BF-Finland
# Take top 10 from each and combine back to one table
#Fin25 <- Fin_BF[1:25,]
#BF25 <- BF_Fin[1:25,]
#BF_Fin_top_sigtab <- rbind(Fin25, BF25)
# Shorten gene names and order again
#BF_Fin_top_sigtab$Gene <- gsub(pattern = "_[A-Z].*", replacement = "", BF_Fin_top_sigtab$Gene)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 10)
#p <- ggplot(data = BF_Fin_top_sigtab,aes(x = Gene, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) + ylab(" Log2 Fold Change\n Burkina Faso Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Enriched ARGs in Burkinabe / Finnish samples (25 with highest log 2 fold change values of each country)") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 10),
# axis.text.y = element_text(family = "Times", size = 12, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold", hjust = -0.01),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 20),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 20, face = "bold"))
#ggsave(filename = "BF_FIN_deseq50_resfinder_genes_new.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1) DESeq2
ResFinder with clusters
#OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
#match <- match(rownames(metadata), colnames(OTU_resfinder))
#OTU_resfinder <- OTU_resfinder[,match]
#all(colnames(OTU_resfinder) == rownames(metadata))
# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
#clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
#colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
#col_order <- c("Class", "Cluster_name", "Gene")
#clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
#match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
#clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
#all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)
# Divide by ARG gene lengths
## Get the lengths in terminal
# seqkit fx2tab --length --name --header-line resfinder.fasta > resfinder_lengths.txt
#resfinder_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/resfinder_lengths.txt", header=FALSE, comment.char="#")
#all(rownames(clusters_tax_table_resfinder$Gene) == resfinder_lengths$V1)
#OTU_resfinder_length_norm <- OTU_resfinder/resfinder_lengths[, 2]
# Normalization with Metaxa2 SSU counts
#deseq_OTU_resfinder <- t(t(OTU_resfinder_length_norm)/metadata$SSU_counts) * 1540
#all(rownames(metadata) == colnames(deseq_OTU_resfinder))
#identical(OTU_resfinder_length_norm[2025, 5]/metadata$SSU_counts[5], deseq_OTU_resfinder[2025, 5])
#all(rownames(OTU_resfinder_length_norm) == clusters_tax_table_resfinder$Gene)
# Deseq
#deseq_OTU <- deseq_OTU_resfinder[, ] * 10^5 + 1
# Merge ARGs so that cluster names instead of gene names can be used
#clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt")
#all(clusters_tax_table_resfinder$Gene == rownames(deseq_OTU))
#Class <- as.data.frame(clusters_tax_table_resfinder$Class)
#Cluster_name <- as.data.frame(clusters_tax_table_resfinder$Cluster_name)
#Gene <- as.data.frame(clusters_tax_table_resfinder$Gene)
#OTU_DESeq <- cbind(deseq_OTU, Class, Cluster_name, Gene)
#names(OTU_DESeq)[names(OTU_DESeq) == "clusters_tax_table_resfinder$Class"] <- "Class"
#names(OTU_DESeq)[names(OTU_DESeq) == "clusters_tax_table_resfinder$Cluster_name"] <- "Cluster_name"
#names(OTU_DESeq)[names(OTU_DESeq) == "clusters_tax_table_resfinder$Gene"] <- "Gene"
#all(OTU_DESeq$Gene == clusters_tax_table_resfinder$Gene)
# Merge
#merged_deseq <- ddply(OTU_DESeq,"Cluster_name",numcolwise(sum))
# Remove duplicates
#uniq_clust <- clusters_tax_table_resfinder[!duplicated(clusters_tax_table_resfinder[ , c("Cluster_name")]),]
# Reorder to match OTU
#match <- match(uniq_clust$Cluster_name, merged_deseq$Cluster_name)
#merged_deseq <- merged_deseq[match,]
#all(merged_deseq$Cluster_name == uniq_clust$Cluster_name)
# Tax table
#match <- match(merged_deseq$Cluster_name, clusters_tax_table_resfinder$Cluster_name)
#heat_clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
#identical(merged_deseq$Cluster_name, heat_clusters_tax_table_resfinder$Cluster_name)
# Reload to get new row numbers
#write.table(merged_deseq, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/merged_deseq.txt",
# row.names=T, sep = "\t", col.names = T)
#merged_deseq <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/merged_deseq.txt", row.names=NULL)
#merged_deseq$row.names<-NULL
#merged_deseq$Cluster_name <- NULL
# Reload to get new row numbers
#write.table(heat_clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
#heat_clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_clusters_tax_table_resfinder.txt", row.names=NULL)
# Create phyloseq
#merged_deseq_resfinder_PHY <- phyloseq(otu_table(merged_deseq, taxa_are_rows = TRUE), sample_data(metadata),
# tax_table(as.matrix(heat_clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
#merged_deseq_resfinder_PHY = subset_samples(merged_deseq_resfinder_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
#merged_deseq_resfinder_PHY = subset_samples(merged_deseq_resfinder_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
#merged_deseq_resfinder_PHY <- subset_samples(merged_deseq_resfinder_PHY, country != "UK")
## Exclude biological / technical replicates
#merged_deseq_resfinder_PHY_stat <- subset_samples(merged_deseq_resfinder_PHY, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10" & alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
#merged_deseq_resfinder_PHY_stat <- subset_samples(merged_deseq_resfinder_PHY_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
# Take pair wise comparisons
#deseq_PHY = subset_samples(merged_deseq_resfinder_PHY_stat, country == "Benin" | country == "Finland")
#deseq_PHY = subset_samples(merged_deseq_resfinder_PHY_stat, country == "Burkina Faso" | country == "Finland")
#deseq_PHY = subset_samples(merged_deseq_resfinder_PHY_stat, country == "Benin" | country == "Burkina Faso")
#hist(log10(apply(otu_table(deseq_PHY), 1, var)), xlab = "log10(variance)")
#varianceThreshold = 50
#keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
#deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
#deseq_PHY
#dds = phyloseq_to_deseq2(deseq_PHY, ~country)
#dds$category <- relevel(dds$country, "Benin", "Finland")
#dds$category <- relevel(dds$country, "Burkina Faso", "Finland")
#dds$category <- relevel(dds$country, "Benin", "Burkina Faso")
#dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)
#res = results(dds, cooksCutoff = FALSE, alpha = 0.05)
#resultsNames(dds)
#plotDispEsts(dds)
#head(res)
#summary(res)
#res = res[order(res$padj, na.last=NA), ]
#alpha = 0.05
#sigtab_resfinder = res[which(res$padj < alpha), ]
#sigtab_resfinder = cbind(as(sigtab_resfinder, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_resfinder), ], "matrix"))
#otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
#otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
#n <- rowSums(otu_table(deseq_PHY))
#sigtab_resfinder = merge(sigtab_resfinder, as.data.frame(n), by = 0)
#kable(sigtab_resfinder, caption = "Taxonomy")
# Plot
# Class
#x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Class, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_resfinder$Class = factor(as.character(sigtab_resfinder$Class), levels = names(x))
# Gene cluster
#x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Cluster_name, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_resfinder$Cluster_name = factor(as.character(sigtab_resfinder$Cluster_name), levels = names(x))
# Gene
#x = tapply(sigtab_resfinder$log2FoldChange, sigtab_resfinder$Gene, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_resfinder$Gene = factor(as.character(sigtab_resfinder$Gene), levels = names(x))
#sorted_sigtab <- sigtab_resfinder[order(-sigtab_resfinder$log2FoldChange), ]DESeq2
Taxonomy (Metaxa2), Genus
metaxa_deseq_PHY <- prune_taxa(taxa_sums(metaxa_PHY_stat) > 0, metaxa_PHY_stat)
metaxa_deseq_PHY <- tax_glom(metaxa_deseq_PHY, taxrank = "Genus")
# Take pair wise comparisons
deseq_PHY = subset_samples(metaxa_deseq_PHY, country == "Benin" | country == "Finland")
# Let's set a threashold for the variance
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 361 taxa and 33 samples ]
## sample_data() Sample Data: [ 33 samples by 32 sample variables ]
## tax_table() Taxonomy Table: [ 361 taxa by 6 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer modedds$category <- relevel(dds$country, "Benin", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## -- replacing outliers and refitting for 121 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## fitting model and testingres = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Benin"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_metaxa = res[which(res$padj < alpha), ]
sigtab_metaxa = cbind(as(sigtab_metaxa, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaxa),
], "matrix"))
n <- rowSums(otu_table(deseq_PHY))
sigtab_metaxa = merge(sigtab_metaxa, as.data.frame(n), by = 0)
kable(sigtab_metaxa, caption = "Taxonomy")| Row.names | baseMean | log2FoldChange | lfcSE | stat | pvalue | padj | Domain | Phylum | Class | Order | Family | Genus | n |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| sp1001 | 5.6367404 | -5.2712840 | 1.2305631 | -4.283636 | 0.0000184 | 0.0000581 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Leuconostocaceae | Weissella | 222 |
| sp1002 | 226.4711573 | 7.1474658 | 0.7521390 | 9.502852 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Streptococcaceae | Lactococcus | 6649 |
| sp1004 | 121.3485273 | 2.5869903 | 0.6157557 | 4.201326 | 0.0000265 | 0.0000789 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Streptococcaceae | Streptococcus | 6203 |
| sp1011 | 23.7862612 | -3.5573650 | 0.6089308 | -5.841986 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Christensenellaceae | Christensenella | 785 |
| sp1014 | 19.3952929 | -5.2836369 | 0.6395652 | -8.261295 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Clostridiaceae | Alkaliphilus | 685 |
| sp1023 | 88.8120991 | -1.2414709 | 0.4912095 | -2.527376 | 0.0114919 | 0.0195145 | Bacteria | Firmicutes | Clostridia | Clostridiales | Clostridiaceae | Clostridium | 3336 |
| sp1033 | 40.5778336 | -2.6098304 | 0.5042277 | -5.175896 | 0.0000002 | 0.0000009 | Bacteria | Firmicutes | Clostridia | Clostridiales | Clostridiaceae | Unclassified Clostridiaceae | 1387 |
| sp1034 | 6.8806767 | -3.8388848 | 0.9577443 | -4.008256 | 0.0000612 | 0.0001694 | Bacteria | Firmicutes | Clostridia | Clostridiales | Eubacteriaceae | Acetobacterium | 252 |
| sp1037 | 54.4844427 | -5.1298583 | 0.9064410 | -5.659341 | 0.0000000 | 0.0000001 | Bacteria | Firmicutes | Clostridia | Clostridiales | Eubacteriaceae | Eubacterium | 1926 |
| sp1040 | 47.5004474 | -5.0241616 | 0.9386956 | -5.352280 | 0.0000001 | 0.0000004 | Bacteria | Firmicutes | Clostridia | Clostridiales | Eubacteriaceae | Unclassified Eubacteriaceae | 1712 |
| sp1051 | 7.7258826 | -2.7524301 | 0.7561909 | -3.639861 | 0.0002728 | 0.0006295 | Bacteria | Firmicutes | Clostridia | Clostridiales | Family XI Incertae Sedis | Sedimentibacter | 256 |
| sp1060 | 15.7416882 | -3.7042979 | 0.6625162 | -5.591256 | 0.0000000 | 0.0000001 | Bacteria | Firmicutes | Clostridia | Clostridiales | Family XIII Incertae Sedis | Anaerovorax | 551 |
| sp1061 | 9.0771710 | -2.0835115 | 0.5815414 | -3.582740 | 0.0003400 | 0.0007650 | Bacteria | Firmicutes | Clostridia | Clostridiales | Family XIII Incertae Sedis | Incertae Sedis | 335 |
| sp1064 | 66.5116507 | -3.3134158 | 0.5535368 | -5.985900 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Family XIII Incertae Sedis | Unclassified Family XIII Incertae Sedis | 2278 |
| sp1085 | 50.6899728 | 4.1404725 | 0.7746020 | 5.345290 | 0.0000001 | 0.0000004 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Anaerostipes | 2820 |
| sp1086 | 173.8037531 | 3.3260076 | 0.7842696 | 4.240898 | 0.0000223 | 0.0000673 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Blautia | 8902 |
| sp1091 | 8.3321466 | 1.8924213 | 0.7006331 | 2.701016 | 0.0069128 | 0.0123198 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Coprococcus | 396 |
| sp1092 | 11.6389506 | 2.9709961 | 0.8618815 | 3.447105 | 0.0005666 | 0.0011999 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Dorea | 572 |
| sp1097 | 183.5231152 | 1.9977014 | 0.5446081 | 3.668145 | 0.0002443 | 0.0005764 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Incertae Sedis | 8805 |
| sp1099 | 3.9087134 | 2.9932455 | 0.7773562 | 3.850546 | 0.0001179 | 0.0003074 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Lachnospira | 207 |
| sp1104 | 33.8018059 | 2.3208395 | 0.7667160 | 3.026987 | 0.0024700 | 0.0046556 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Pseudobutyrivibrio | 1668 |
| sp1106 | 61.6616549 | 2.7680705 | 0.8268654 | 3.347668 | 0.0008149 | 0.0016575 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Roseburia | 3194 |
| sp1110 | 598.2247957 | 2.4010220 | 0.6509168 | 3.688677 | 0.0002254 | 0.0005374 | Bacteria | Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Unclassified Lachnospiraceae | 29461 |
| sp1114 | 7.4357280 | -6.3817156 | 1.2147131 | -5.253681 | 0.0000001 | 0.0000006 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptococcaceae | Dehalobacterium | 206 |
| sp1117 | 9.0908661 | -5.0583684 | 0.8666852 | -5.836454 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptococcaceae | Desulfitobacterium | 258 |
| sp1120 | 9.2927562 | -2.7520313 | 0.6309079 | -4.362018 | 0.0000129 | 0.0000426 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptococcaceae | Desulfotomaculum | 285 |
| sp1128 | 36.7606214 | -1.8844007 | 0.5138135 | -3.667480 | 0.0002450 | 0.0005764 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptococcaceae | Unclassified Peptococcaceae | 1209 |
| sp1130 | 110.5787088 | -1.4183080 | 0.4681166 | -3.029818 | 0.0024470 | 0.0046364 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptostreptococcaceae | Incertae Sedis | 4042 |
| sp1132 | 15.3447920 | -2.3421447 | 1.0662288 | -2.196662 | 0.0280446 | 0.0429619 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptostreptococcaceae | Proteocatella | 768 |
| sp1135 | 11.5345084 | -2.2542023 | 0.5904532 | -3.817749 | 0.0001347 | 0.0003430 | Bacteria | Firmicutes | Clostridia | Clostridiales | Peptostreptococcaceae | Unclassified Peptostreptococcaceae | 428 |
| sp1138 | 14.8027870 | -3.5455714 | 0.8647704 | -4.100015 | 0.0000413 | 0.0001162 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Anaerofilum | 494 |
| sp1142 | 258.7057088 | 2.3968763 | 0.6945092 | 3.451180 | 0.0005581 | 0.0011960 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Faecalibacterium | 12558 |
| sp1143 | 8.6276203 | -2.7776240 | 0.8522525 | -3.259156 | 0.0011174 | 0.0022103 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Fastidiosipila | 268 |
| sp1147 | 34.1427172 | 1.3399962 | 0.5814816 | 2.304451 | 0.0211973 | 0.0333822 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Oscillibacter | 1492 |
| sp1152 | 34.2119628 | -5.9933949 | 0.9123862 | -6.568923 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Saccharofermentans | 1114 |
| sp1154 | 59.5891723 | 2.9811885 | 0.5157615 | 5.780168 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Subdoligranulum | 2988 |
| sp1159 | 49.5333151 | -7.1291543 | 1.1050675 | -6.451329 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Syntrophomonadaceae | Syntrophomonas | 1632 |
| sp1162 | 22.6739643 | -7.6235041 | 1.2124788 | -6.287536 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Syntrophomonadaceae | Unclassified Syntrophomonadaceae | 752 |
| sp1168 | 22.9964629 | -6.9331196 | 1.1779873 | -5.885564 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Anaeroarcus | 523 |
| sp1170 | 26.3485944 | -6.8445570 | 1.1939242 | -5.732824 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Anaerosinus | 1165 |
| sp1175 | 23.6568502 | 2.4640759 | 0.7763764 | 3.173816 | 0.0015045 | 0.0029277 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Dialister | 1177 |
| sp1176 | 9.0948215 | -4.4835111 | 1.1111309 | -4.035088 | 0.0000546 | 0.0001523 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Megamonas | 1052 |
| sp1181 | 23.1423650 | -5.3257792 | 0.6588084 | -8.083957 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Pelosinus | 904 |
| sp1188 | 62.4900316 | -4.6354653 | 0.5891797 | -7.867659 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Sporomusa | 1650 |
| sp1190 | 5.4027320 | -5.9344536 | 1.4073563 | -4.216738 | 0.0000248 | 0.0000744 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Succinispira | 208 |
| sp1191 | 11.6432939 | -5.6788473 | 0.9013412 | -6.300441 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Thermosinus | 335 |
| sp1192 | 387.9338361 | -2.0325332 | 0.5646769 | -3.599462 | 0.0003189 | 0.0007220 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Unclassified Veillonellaceae | 11739 |
| sp1193 | 10.5434990 | -1.5096746 | 0.6340157 | -2.381131 | 0.0172596 | 0.0279885 | Bacteria | Firmicutes | Clostridia | Clostridiales | Veillonellaceae | Veillonella | 525 |
| sp1215 | 9.1870088 | -2.9896777 | 0.6773683 | -4.413666 | 0.0000102 | 0.0000339 | Bacteria | Firmicutes | Clostridia | Thermoanaerobacterales | Family III Incertae Sedis | Thermoanaerobacterium | 302 |
| sp1242 | 15.6587558 | 2.3827015 | 1.0196990 | 2.336672 | 0.0194563 | 0.0309923 | Bacteria | Firmicutes | Erysipelotrichi | Erysipelotrichales | Erysipelotrichaceae | Catenibacterium | 850 |
| sp1254 | 1.7805128 | -4.3453883 | 1.2153137 | -3.575528 | 0.0003495 | 0.0007741 | Bacteria | Fusobacteria | Fusobacteria | Fusobacteriales | Fusobacteriaceae | Cetobacterium | 238 |
| sp1255 | 5.4853700 | -3.2582528 | 0.9697767 | -3.359797 | 0.0007800 | 0.0015955 | Bacteria | Fusobacteria | Fusobacteria | Fusobacteriales | Fusobacteriaceae | Fusobacterium | 969 |
| sp1259 | 8.0319761 | -6.1296162 | 1.0709803 | -5.723370 | 0.0000000 | 0.0000000 | Bacteria | Fusobacteria | Fusobacteria | Fusobacteriales | Fusobacteriaceae | Unclassified Fusobacteriaceae | 877 |
| sp1261 | 2.1270707 | -3.6184498 | 1.0736812 | -3.370134 | 0.0007513 | 0.0015605 | Bacteria | Fusobacteria | Fusobacteria | Fusobacteriales | Leptotrichiaceae | Sebaldella | 137 |
| sp1274 | 24.9264841 | -3.2560298 | 0.7907876 | -4.117452 | 0.0000383 | 0.0001095 | Bacteria | Lentisphaerae | Lentisphaeria | Victivallales | Victivallaceae | Victivallis | 733 |
| sp1304 | 19.2206239 | -5.6042506 | 0.7925562 | -7.071108 | 0.0000000 | 0.0000000 | Bacteria | Planctomycetes | Planctomycetacia | Planctomycetales | Planctomycetaceae | Unclassified Planctomycetaceae | 643 |
| sp1311 | 7.7832179 | 1.4332742 | 0.6409647 | 2.236120 | 0.0253439 | 0.0389906 | Bacteria | Proteobacteria | Alphaproteobacteria | Caulobacterales | Caulobacteraceae | Brevundimonas | 376 |
| sp1353 | 3.2161858 | -4.1164212 | 1.5435609 | -2.666834 | 0.0076569 | 0.0134145 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhizobiales | Bradyrhizobiaceae | Rhodopseudomonas | 504 |
| sp1354 | 7.9692023 | -2.0761161 | 0.7027508 | -2.954271 | 0.0031341 | 0.0058158 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhizobiales | Bradyrhizobiaceae | Unclassified Bradyrhizobiaceae | 947 |
| sp1388 | 7.2608632 | -4.1994518 | 1.3499285 | -3.110870 | 0.0018654 | 0.0035720 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhizobiales | Methylocystaceae | Pleomorphomonas | 236 |
| sp1389 | 5.4508317 | -3.9838397 | 1.1925073 | -3.340726 | 0.0008356 | 0.0016900 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhizobiales | Methylocystaceae | Unclassified Methylocystaceae | 147 |
| sp141 | 33.2875347 | -7.4092762 | 1.2020823 | -6.163701 | 0.0000000 | 0.0000000 | Bacteria | Acidobacteria | Holophagae | Holophagales | Holophagaceae | Geothrix | 776 |
| sp142 | 10.3184136 | -6.1153904 | 1.2304022 | -4.970237 | 0.0000007 | 0.0000024 | Bacteria | Acidobacteria | Holophagae | Holophagales | Holophagaceae | Holophaga | 270 |
| sp1422 | 6.4334199 | -2.8803648 | 1.2836277 | -2.243926 | 0.0248372 | 0.0383751 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhizobiales | Xanthobacteraceae | Xanthobacter | 161 |
| sp143 | 16.1507622 | -6.7641998 | 1.2992501 | -5.206234 | 0.0000002 | 0.0000008 | Bacteria | Acidobacteria | Holophagae | Holophagales | Holophagaceae | Unclassified Holophagaceae | 423 |
| sp1443 | 25.1501535 | 4.1959036 | 0.6214193 | 6.752129 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodobacterales | Rhodobacteraceae | Paracoccus | 1315 |
| sp1494 | 6.1679407 | -1.6434848 | 0.6164621 | -2.665995 | 0.0076761 | 0.0134145 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodospirillales | Acetobacteraceae | Unclassified Acetobacteraceae | 283 |
| sp1503 | 14.5591082 | -2.5079297 | 0.9118086 | -2.750500 | 0.0059504 | 0.0107108 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodospirillales | Rhodospirillaceae | Azospirillum | 423 |
| sp1511 | 107.9050636 | -3.1078868 | 1.0786276 | -2.881334 | 0.0039600 | 0.0073107 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodospirillales | Rhodospirillaceae | Magnetospirillum | 1634 |
| sp1524 | 2.6147227 | -4.1092344 | 1.9030420 | -2.159298 | 0.0308271 | 0.0468259 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodospirillales | Rhodospirillaceae | Telmatospirillum | 394 |
| sp1529 | 145.6101123 | -3.2894418 | 0.8908697 | -3.692394 | 0.0002222 | 0.0005367 | Bacteria | Proteobacteria | Alphaproteobacteria | Rhodospirillales | Rhodospirillaceae | Unclassified Rhodospirillaceae | 2031 |
| sp1595 | 21.3772433 | -2.5214428 | 0.7911840 | -3.186923 | 0.0014379 | 0.0028134 | Bacteria | Proteobacteria | Alphaproteobacteria | Sphingomonadales | Sphingomonadaceae | Unclassified Sphingomonadaceae | 843 |
| sp161 | 2.9103521 | 4.0271922 | 1.3454453 | 2.993204 | 0.0027607 | 0.0051762 | Bacteria | Actinobacteria | Actinobacteria | Actinomycetales | Actinomycetaceae | Actinobaculum | 197 |
| sp1615 | 47.2976209 | -1.1976978 | 0.5198829 | -2.303784 | 0.0212348 | 0.0333822 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Alcaligenaceae | Unclassified Alcaligenaceae | 20639 |
| sp1616 | 49.9523100 | -2.8837274 | 0.6902914 | -4.177551 | 0.0000295 | 0.0000862 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Burkholderiaceae | Burkholderia | 2317 |
| sp1620 | 42.6927060 | -2.6048315 | 0.6591266 | -3.951944 | 0.0000775 | 0.0002083 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Burkholderiaceae | Cupriavidus | 2993 |
| sp1621 | 4.7773790 | -2.9529267 | 1.3495402 | -2.188098 | 0.0286624 | 0.0437224 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Burkholderiaceae | Limnobacter | 961 |
| sp1624 | 9.4760492 | -3.2055465 | 0.9526088 | -3.365019 | 0.0007654 | 0.0015745 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Burkholderiaceae | Ralstonia | 393 |
| sp1625 | 21.8075771 | -3.6458660 | 0.7070027 | -5.156792 | 0.0000003 | 0.0000010 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Burkholderiaceae | Unclassified Burkholderiaceae | 883 |
| sp1629 | 13.5847529 | -2.0820992 | 0.6180055 | -3.369063 | 0.0007542 | 0.0015605 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Aquabacterium | 485 |
| sp1632 | 11.2768188 | -2.0464261 | 0.5346484 | -3.827611 | 0.0001294 | 0.0003327 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Brachymonas | 434 |
| sp1636 | 274.8555174 | 2.8703755 | 0.5066583 | 5.665308 | 0.0000000 | 0.0000001 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Comamonas | 13242 |
| sp1639 | 3.9584467 | -5.1283109 | 1.2952598 | -3.959291 | 0.0000752 | 0.0002050 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Diaphorobacter | 136 |
| sp1641 | 75.5059585 | -7.0200830 | 1.1839110 | -5.929570 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Hydrogenophaga | 2580 |
| sp1648 | 48.0902650 | -8.7040751 | 1.5683988 | -5.549657 | 0.0000000 | 0.0000001 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Macromonas | 1495 |
| sp1649 | 9.9339760 | -6.7952855 | 1.5943778 | -4.262030 | 0.0000203 | 0.0000629 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Malikia | 327 |
| sp166 | 2.2473810 | 2.4667878 | 0.9826025 | 2.510464 | 0.0120573 | 0.0201889 | Bacteria | Actinobacteria | Actinobacteria | Actinomycetales | Actinomycetaceae | Unclassified Actinomycetaceae | 134 |
| sp1665 | 1.9047948 | 2.5941385 | 0.9908340 | 2.618136 | 0.0088411 | 0.0152288 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Simplicispira | 89 |
| sp1666 | 7.1755481 | 5.1085150 | 0.9932698 | 5.143129 | 0.0000003 | 0.0000010 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Comamonadaceae | Sphaerotilus | 476 |
| sp1678 | 11.9570083 | -5.3256539 | 1.0366007 | -5.137613 | 0.0000003 | 0.0000010 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Oxalobacteraceae | Herbaspirillum | 353 |
| sp1685 | 47.5122473 | -1.3397200 | 0.5492956 | -2.438978 | 0.0147289 | 0.0241018 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Oxalobacteraceae | Unclassified Oxalobacteraceae | 5867 |
| sp1686 | 4.1204168 | -2.8679316 | 0.8591402 | -3.338141 | 0.0008434 | 0.0016962 | Bacteria | Proteobacteria | Betaproteobacteria | Burkholderiales | Oxalobacteraceae | Undibacterium | 221 |
| sp1692 | 243.0323162 | -7.5026911 | 1.0708314 | -7.006417 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Hydrogenophilales | Hydrogenophilaceae | Thiobacillus | 6401 |
| sp1700 | 8.8916502 | -3.3592862 | 0.9603825 | -3.497863 | 0.0004690 | 0.0010233 | Bacteria | Proteobacteria | Betaproteobacteria | Methylophilales | Methylophilaceae | Unclassified Methylophilaceae | 242 |
| sp1710 | 22.4371642 | -2.9900679 | 0.7019983 | -4.259366 | 0.0000205 | 0.0000631 | Bacteria | Proteobacteria | Betaproteobacteria | Neisseriales | Neisseriaceae | Chromobacterium | 1494 |
| sp1714 | 19.2891809 | -4.7181739 | 1.2536884 | -3.763434 | 0.0001676 | 0.0004112 | Bacteria | Proteobacteria | Betaproteobacteria | Neisseriales | Neisseriaceae | Formivibrio | 704 |
| sp1719 | 106.8944273 | -3.9443575 | 0.9176064 | -4.298529 | 0.0000172 | 0.0000559 | Bacteria | Proteobacteria | Betaproteobacteria | Neisseriales | Neisseriaceae | Laribacter | 3884 |
| sp1728 | 233.7356886 | -2.0373357 | 0.5797357 | -3.514249 | 0.0004410 | 0.0009680 | Bacteria | Proteobacteria | Betaproteobacteria | Neisseriales | Neisseriaceae | Unclassified Neisseriaceae | 9236 |
| sp173 | 86.2347517 | 3.1241906 | 0.7308409 | 4.274789 | 0.0000191 | 0.0000599 | Bacteria | Actinobacteria | Actinobacteria | Bifidobacteriales | Bifidobacteriaceae | Bifidobacterium | 4668 |
| sp1739 | 7.2982580 | -3.5475786 | 1.0742995 | -3.302225 | 0.0009592 | 0.0019078 | Bacteria | Proteobacteria | Betaproteobacteria | Nitrosomonadales | Nitrosomonadaceae | Nitrosomonas | 222 |
| sp1745 | 46.4032828 | -4.2192420 | 0.6922577 | -6.094900 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | 12up | 7066 |
| sp1747 | 3.1461849 | -2.3983286 | 1.0498202 | -2.284514 | 0.0223414 | 0.0349691 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Azonexus | 105 |
| sp1749 | 1.7506507 | -3.9715360 | 1.6077685 | -2.470216 | 0.0135031 | 0.0221969 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Azovibrio | 363 |
| sp1750 | 576.5182166 | -5.3052076 | 0.7591859 | -6.988022 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Dechloromonas | 29563 |
| sp1754 | 199.8273580 | -5.3475284 | 0.6726205 | -7.950291 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Propionivibrio | 6606 |
| sp1756 | 21.1143446 | -6.6967650 | 1.3319016 | -5.027973 | 0.0000005 | 0.0000018 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Rhodocyclus | 313 |
| sp1758 | 76.4730139 | 1.6109328 | 0.5255192 | 3.065412 | 0.0021737 | 0.0041404 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Thauera | 5384 |
| sp1760 | 1800.5104561 | -4.0864690 | 0.6246441 | -6.542076 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Betaproteobacteria | Rhodocyclales | Rhodocyclaceae | Unclassified Rhodocyclaceae | 53377 |
| sp1785 | 8.1105559 | -6.5040529 | 1.1506420 | -5.652542 | 0.0000000 | 0.0000001 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfobacterales | Desulfobacteraceae | Desulfobacterium | 226 |
| sp1793 | 26.0367762 | -6.8036043 | 1.0948757 | -6.214042 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfobacterales | Desulfobacteraceae | Desulforegula | 814 |
| sp1798 | 53.7234230 | -8.5046035 | 1.1338236 | -7.500817 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfobacterales | Desulfobacteraceae | Unclassified Desulfobacteraceae | 1612 |
| sp1799 | 196.3551924 | -6.0715494 | 0.6294236 | -9.646206 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfobacterales | Desulfobulbaceae | Desulfobulbus | 6250 |
| sp1806 | 26.5804765 | -6.7554412 | 0.8821625 | -7.657820 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfobacterales | Desulfobulbaceae | Unclassified Desulfobulbaceae | 735 |
| sp1819 | 118.4652921 | -6.5749936 | 1.0107367 | -6.505150 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfovibrionales | Desulfomicrobiaceae | Desulfomicrobium | 8314 |
| sp1825 | 673.9717359 | -4.4773079 | 0.5975181 | -7.493176 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfovibrionales | Desulfovibrionaceae | Desulfovibrio | 18265 |
| sp1827 | 34.1292569 | -2.4071728 | 0.6571722 | -3.662926 | 0.0002494 | 0.0005791 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfovibrionales | Desulfovibrionaceae | Unclassified Desulfovibrionaceae | 988 |
| sp183 | 22.5909245 | -2.2281659 | 0.7518352 | -2.963636 | 0.0030403 | 0.0056710 | Bacteria | Actinobacteria | Actinobacteria | Corynebacteriales | Corynebacteriaceae | Corynebacterium | 639 |
| sp1833 | 40.4675358 | -8.0969201 | 1.4034686 | -5.769221 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfuromonadales | BVA18 | Geobacter | 1644 |
| sp1838 | 29.8828802 | -7.6638465 | 1.0928674 | -7.012604 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfuromonadales | Desulfuromonadaceae | Pelobacter | 1044 |
| sp1844 | 317.6002835 | -7.7523446 | 0.7472837 | -10.374031 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfuromonadales | Geobacteraceae | Geobacter | 10828 |
| sp1847 | 3.9247673 | -4.3871763 | 0.9245929 | -4.744982 | 0.0000021 | 0.0000073 | Bacteria | Proteobacteria | Deltaproteobacteria | Desulfuromonadales | Geobacteraceae | Unclassified Geobacteraceae | 211 |
| sp1883 | 6.4184719 | -6.1463240 | 1.0482095 | -5.863641 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Order Incertae Sedis | Syntrophorhabdaceae | Syntrophorhabdus | 197 |
| sp1887 | 5.4736650 | -5.5675912 | 1.2997003 | -4.283750 | 0.0000184 | 0.0000581 | Bacteria | Proteobacteria | Deltaproteobacteria | Syntrophobacterales | Syntrophaceae | Desulfomonile | 335 |
| sp1888 | 15.6575059 | -7.4473500 | 1.1854232 | -6.282440 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Syntrophobacterales | Syntrophaceae | Smithella | 389 |
| sp189 | 26.5802506 | -2.6265314 | 0.6975261 | -3.765496 | 0.0001662 | 0.0004112 | Bacteria | Actinobacteria | Actinobacteria | Corynebacteriales | Mycobacteriaceae | Mycobacterium | 822 |
| sp1890 | 40.0461581 | -7.7530741 | 1.0843607 | -7.149903 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Deltaproteobacteria | Syntrophobacterales | Syntrophaceae | Unclassified Syntrophaceae | 1029 |
| sp1896 | 3.3336767 | -4.1079246 | 1.2706087 | -3.233037 | 0.0012248 | 0.0024095 | Bacteria | Proteobacteria | Deltaproteobacteria | Syntrophobacterales | Syntrophobacteraceae | Syntrophobacter | 359 |
| sp1898 | 10.1322293 | -5.4640072 | 1.2713186 | -4.297905 | 0.0000172 | 0.0000559 | Bacteria | Proteobacteria | Deltaproteobacteria | Syntrophobacterales | Syntrophobacteraceae | Unclassified Syntrophobacteraceae | 1498 |
| sp1902 | 3.8670638 | -3.9086696 | 0.9204647 | -4.246409 | 0.0000217 | 0.0000663 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Campylobacteraceae | Campylobacter | 180 |
| sp1903 | 1676.0288499 | -7.8675650 | 0.8927338 | -8.812890 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Campylobacteraceae | Sulfurospirillum | 48230 |
| sp1904 | 26.0558618 | -1.4597451 | 0.4231971 | -3.449327 | 0.0005620 | 0.0011971 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Campylobacteraceae | Unclassified Campylobacteraceae | 1351 |
| sp1905 | 12.0287558 | -3.8436244 | 0.6853013 | -5.608663 | 0.0000000 | 0.0000001 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Helicobacteraceae | Helicobacter | 355 |
| sp1906 | 117.5084172 | -7.9623889 | 1.1291554 | -7.051632 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Helicobacteraceae | Sulfuricurvum | 2740 |
| sp1907 | 14.0330662 | -6.2162021 | 1.2174934 | -5.105738 | 0.0000003 | 0.0000012 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Helicobacteraceae | Sulfurimonas | 1295 |
| sp1909 | 76.7105014 | -6.5389512 | 0.8972770 | -7.287551 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Epsilonproteobacteria | Campylobacterales | Helicobacteraceae | Unclassified Helicobacteraceae | 2008 |
| sp1914 | 18.1241701 | -2.1393371 | 0.7940981 | -2.694046 | 0.0070590 | 0.0125185 | Bacteria | Proteobacteria | Epsilonproteobacteria | Nautiliales | Nautiliaceae | Caminibacter | 597 |
| sp1928 | 1069.2231454 | 2.6638689 | 0.7441281 | 3.579853 | 0.0003438 | 0.0007687 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Aeromonadaceae | Aeromonas | 102913 |
| sp1931 | 782.2235133 | -3.8566171 | 0.8675629 | -4.445346 | 0.0000088 | 0.0000295 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Aeromonadaceae | Tolumonas | 53178 |
| sp1932 | 262.7223942 | -2.4013432 | 0.7231776 | -3.320544 | 0.0008984 | 0.0017968 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Aeromonadaceae | Unclassified Aeromonadaceae | 20046 |
| sp1933 | 0.9026186 | -3.3737513 | 1.3636066 | -2.474138 | 0.0133558 | 0.0220554 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Aeromonadaceae | Zobellella | 69 |
| sp1934 | 10.1898579 | -4.6832637 | 1.2651844 | -3.701645 | 0.0002142 | 0.0005210 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Succinivibrionaceae | Anaerobiospirillum | 374 |
| sp1937 | 20.4976080 | -3.7493642 | 1.0324580 | -3.631493 | 0.0002818 | 0.0006461 | Bacteria | Proteobacteria | Gammaproteobacteria | Aeromonadales | Succinivibrionaceae | Unclassified Succinivibrionaceae | 788 |
| sp1989 | 11.5390076 | 1.8013880 | 0.7212123 | 2.497722 | 0.0124994 | 0.0208005 | Bacteria | Proteobacteria | Gammaproteobacteria | Chromatiales | Chromatiaceae | Rheinheimera | 478 |
| sp2001 | 4.7955592 | -3.2716757 | 0.9382783 | -3.486893 | 0.0004887 | 0.0010534 | Bacteria | Proteobacteria | Gammaproteobacteria | Chromatiales | Chromatiaceae | Unclassified Chromatiaceae | 1758 |
| sp2012 | 1.9083375 | -3.4014311 | 1.2706769 | -2.676865 | 0.0074314 | 0.0131143 | Bacteria | Proteobacteria | Gammaproteobacteria | Chromatiales | Ectothiorhodospiraceae | Unclassified Ectothiorhodospiraceae | 136 |
| sp2016 | 276.4250798 | -8.6317438 | 0.9915256 | -8.705518 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Gammaproteobacteria | Chromatiales | Halothiobacillaceae | Thiovirga | 7600 |
| sp2017 | 8.3377890 | -6.5256703 | 1.2673797 | -5.148947 | 0.0000003 | 0.0000010 | Bacteria | Proteobacteria | Gammaproteobacteria | Chromatiales | Halothiobacillaceae | Unclassified Halothiobacillaceae | 210 |
| sp2034 | 21.7125672 | 3.9783187 | 1.1582868 | 3.434658 | 0.0005933 | 0.0012491 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Citrobacter | 692 |
| sp2036 | 4.9031851 | -2.6509393 | 0.8465626 | -3.131416 | 0.0017397 | 0.0033491 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Dickeya | 378 |
| sp2040 | 199.0416181 | 3.2674963 | 0.6821537 | 4.789971 | 0.0000017 | 0.0000059 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Escherichia-Shigella | 6817 |
| sp2043 | 17.5830659 | 2.2288384 | 0.5839779 | 3.816649 | 0.0001353 | 0.0003430 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Klebsiella | 1111 |
| sp2057 | 1.3746613 | 2.4248437 | 1.0654842 | 2.275814 | 0.0228571 | 0.0356215 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Serratia | 178 |
| sp2065 | 4.2909039 | 3.3419034 | 1.2160611 | 2.748138 | 0.0059935 | 0.0107346 | Bacteria | Proteobacteria | Gammaproteobacteria | Enterobacteriales | Enterobacteriaceae | Yersinia | 150 |
| sp2079 | 2.5126988 | -4.8056208 | 1.2415792 | -3.870571 | 0.0001086 | 0.0002874 | Bacteria | Proteobacteria | Gammaproteobacteria | Methylococcales | Methylococcaceae | Methylocaldum | 164 |
| sp2083 | 5.0961778 | -4.7354331 | 1.2255567 | -3.863904 | 0.0001116 | 0.0002932 | Bacteria | Proteobacteria | Gammaproteobacteria | Methylococcales | Methylococcaceae | Methylomonas | 159 |
| sp2146 | 2.3473753 | -4.7589714 | 1.8887246 | -2.519675 | 0.0117463 | 0.0197602 | Bacteria | Proteobacteria | Gammaproteobacteria | Pasteurellales | Pasteurellaceae | Gallibacterium | 235 |
| sp2153 | 6.6644081 | -2.0844358 | 0.7369736 | -2.828372 | 0.0046785 | 0.0085496 | Bacteria | Proteobacteria | Gammaproteobacteria | Pasteurellales | Pasteurellaceae | Unclassified Pasteurellaceae | 188 |
| sp2155 | 1647.6292223 | 4.0957589 | 0.5507880 | 7.436180 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Gammaproteobacteria | Pseudomonadales | Moraxellaceae | Acinetobacter | 81019 |
| sp2157 | 8.2419352 | 3.1297030 | 0.8317126 | 3.762962 | 0.0001679 | 0.0004112 | Bacteria | Proteobacteria | Gammaproteobacteria | Pseudomonadales | Moraxellaceae | Enhydrobacter | 396 |
| sp2161 | 13.6617519 | 2.2137667 | 0.5528078 | 4.004587 | 0.0000621 | 0.0001707 | Bacteria | Proteobacteria | Gammaproteobacteria | Pseudomonadales | Moraxellaceae | Unclassified Moraxellaceae | 686 |
| sp2164 | 4.4339102 | 4.3941874 | 0.7003819 | 6.273988 | 0.0000000 | 0.0000000 | Bacteria | Proteobacteria | Gammaproteobacteria | Pseudomonadales | Pseudomonadaceae | Cellvibrio | 247 |
| sp2182 | 3.7815702 | -5.4177504 | 1.9334573 | -2.802105 | 0.0050770 | 0.0092310 | Bacteria | Proteobacteria | Gammaproteobacteria | Thiotrichales | Thiotrichaceae | Beggiatoa | 511 |
| sp2191 | 1.1945246 | -3.0337791 | 1.1457039 | -2.647961 | 0.0080979 | 0.0140156 | Bacteria | Proteobacteria | Gammaproteobacteria | Thiotrichales | Thiotrichaceae | Thiothrix | 481 |
| sp2200 | 28.4683370 | 3.7542630 | 0.9056046 | 4.145587 | 0.0000339 | 0.0000976 | Bacteria | Proteobacteria | Gammaproteobacteria | Vibrionales | Vibrionaceae | Unclassified Vibrionaceae | 934 |
| sp2226 | 11.7860338 | 2.1825110 | 0.5516185 | 3.956559 | 0.0000760 | 0.0002058 | Bacteria | Proteobacteria | Gammaproteobacteria | Xanthomonadales | Xanthomonadaceae | Stenotrophomonas | 563 |
| sp2241 | 9.6606973 | -5.4120994 | 0.9976201 | -5.425010 | 0.0000001 | 0.0000002 | Bacteria | Spirochaetes | Spirochaetes | Spirochaetales | Leptospiraceae | Leptospira | 273 |
| sp2243 | 6.9156927 | -6.2775973 | 1.0849557 | -5.786040 | 0.0000000 | 0.0000000 | Bacteria | Spirochaetes | Spirochaetes | Spirochaetales | Leptospiraceae | Unclassified Leptospiraceae | 208 |
| sp2245 | 33.5781231 | -7.8275347 | 1.0872880 | -7.199136 | 0.0000000 | 0.0000000 | Bacteria | Spirochaetes | Spirochaetes | Spirochaetales | Spirochaetaceae | Spirochaeta | 936 |
| sp2247 | 77.7996331 | -5.7417449 | 0.8319653 | -6.901424 | 0.0000000 | 0.0000000 | Bacteria | Spirochaetes | Spirochaetes | Spirochaetales | Spirochaetaceae | Treponema | 2705 |
| sp2248 | 71.1600328 | -6.6501273 | 0.9150171 | -7.267763 | 0.0000000 | 0.0000000 | Bacteria | Spirochaetes | Spirochaetes | Spirochaetales | Spirochaetaceae | Unclassified Spirochaetaceae | 2277 |
| sp2252 | 14.8090329 | -7.3669168 | 0.9791738 | -7.523605 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Aminobacterium | 380 |
| sp2253 | 39.7866619 | -7.6758573 | 1.0349175 | -7.416879 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Aminomonas | 951 |
| sp2255 | 13.4890182 | -7.2427213 | 0.9616861 | -7.531274 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Candidatus Tammella | 392 |
| sp2256 | 67.7520406 | -5.2343121 | 0.8662914 | -6.042207 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Cloacibacillus | 2118 |
| sp2260 | 13.6050617 | -6.8868733 | 0.9183209 | -7.499419 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Thermanaerovibrio | 381 |
| sp2262 | 565.5248601 | -7.4127075 | 0.6663890 | -11.123694 | 0.0000000 | 0.0000000 | Bacteria | Synergistetes | Synergistia | Synergistales | Synergistaceae | Unclassified Synergistaceae | 16391 |
| sp2277 | 1.2538935 | -3.5436648 | 1.1289025 | -3.139035 | 0.0016950 | 0.0032807 | Bacteria | Tenericutes | Mollicutes | Entomoplasmatales | Spiroplasmataceae | Spiroplasma | 92 |
| sp2284 | 3.9518008 | -3.1193017 | 0.8245148 | -3.783197 | 0.0001548 | 0.0003871 | Bacteria | Tenericutes | Mollicutes | Mycoplasmatales | Mycoplasmataceae | Mycoplasma | 212 |
| sp2313 | 53.8448408 | -5.3526868 | 0.8940776 | -5.986826 | 0.0000000 | 0.0000000 | Bacteria | Verrucomicrobia | Opitutae | Opitutales | Opitutaceae | Opitutus | 2139 |
| sp2314 | 11.0414308 | -6.2479414 | 1.0318267 | -6.055224 | 0.0000000 | 0.0000000 | Bacteria | Verrucomicrobia | Opitutae | Opitutales | Opitutaceae | Unclassified Opitutaceae | 448 |
| sp2321 | 4.2900049 | -3.9163496 | 1.0320974 | -3.794554 | 0.0001479 | 0.0003724 | Bacteria | Verrucomicrobia | Opitutae | Puniceicoccales | Puniceicoccaceae | Unclassified Puniceicoccaceae | 156 |
| sp289 | 43.0189522 | 3.3005045 | 0.7899296 | 4.178226 | 0.0000294 | 0.0000862 | Bacteria | Actinobacteria | Actinobacteria | Micrococcales | Microbacteriaceae | Microbacterium | 1782 |
| sp355 | 4.0421918 | 1.6669756 | 0.6676913 | 2.496626 | 0.0125381 | 0.0208005 | Bacteria | Actinobacteria | Actinobacteria | Propionibacteriales | Propionibacteriaceae | Propionibacterium | 193 |
| sp357 | 16.3666367 | -2.2914041 | 0.9710826 | -2.359639 | 0.0182927 | 0.0295309 | Bacteria | Actinobacteria | Actinobacteria | Propionibacteriales | Propionibacteriaceae | Propioniciclava | 609 |
| sp360 | 7.1701386 | 3.3509210 | 0.6720612 | 4.986035 | 0.0000006 | 0.0000022 | Bacteria | Actinobacteria | Actinobacteria | Propionibacteriales | Propionibacteriaceae | Tessaracoccus | 383 |
| sp382 | 5.9053697 | -1.2076194 | 0.5064720 | -2.384375 | 0.0171081 | 0.0278685 | Bacteria | Actinobacteria | Actinobacteria | Streptomycetales | Streptomycetaceae | Streptomyces | 226 |
| sp410 | 34.0006684 | 2.9901330 | 0.7203662 | 4.150851 | 0.0000331 | 0.0000962 | Bacteria | Actinobacteria | Coriobacteriia | Coriobacteriales | Coriobacteriaceae | Collinsella | 1681 |
| sp420 | 24.7722449 | 1.8890539 | 0.4933146 | 3.829308 | 0.0001285 | 0.0003327 | Bacteria | Actinobacteria | Coriobacteriia | Coriobacteriales | Coriobacteriaceae | Unclassified Coriobacteriaceae | 1103 |
| sp466 | 750.0975359 | 0.8691634 | 0.3839159 | 2.263942 | 0.0235777 | 0.0365860 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Bacteroidaceae | Bacteroides | 31191 |
| sp467 | 10.3346673 | -3.2471820 | 0.6884622 | -4.716573 | 0.0000024 | 0.0000083 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Bacteroidaceae | Unclassified Bacteroidaceae | 293 |
| sp470 | 12.3172639 | -4.4507603 | 0.7101646 | -6.267223 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Marinilabiaceae | Alkaliflexus | 442 |
| sp472 | 3.7255048 | -4.3783999 | 1.0675483 | -4.101360 | 0.0000411 | 0.0001162 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Marinilabiaceae | Cytophaga | 146 |
| sp473 | 9.9482299 | -5.7391180 | 0.9629606 | -5.959868 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Marinilabiaceae | Marinifilum | 305 |
| sp475 | 16.9502961 | -4.9810338 | 0.7012686 | -7.102890 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Marinilabiaceae | Unclassified Marinilabiaceae | 557 |
| sp480 | 7.2354819 | -5.9978567 | 1.1398684 | -5.261885 | 0.0000001 | 0.0000006 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Candidatus Armantifilum | 269 |
| sp483 | 13.0354989 | 1.9079301 | 0.6717795 | 2.840114 | 0.0045097 | 0.0082832 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Odoribacter | 568 |
| sp484 | 156.4649474 | -3.9589686 | 0.5772266 | -6.858604 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Paludibacter | 5308 |
| sp485 | 731.9283190 | -2.2727140 | 0.6203675 | -3.663496 | 0.0002488 | 0.0005791 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Parabacteroides | 24997 |
| sp486 | 6.0813348 | -2.4667944 | 0.9437911 | -2.613708 | 0.0089566 | 0.0153541 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Petrimonas | 177 |
| sp487 | 21.0115353 | -4.0961176 | 0.6276076 | -6.526559 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Porphyromonas | 647 |
| sp490 | 371.3555332 | -3.1632297 | 0.5028316 | -6.290833 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Unclassified Porphyromonadaceae | 12293 |
| sp498 | 9.9795994 | -6.4535406 | 1.3738451 | -4.697430 | 0.0000026 | 0.0000090 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Rikenellaceae | Blvii28 wastewater-sludge group | 764 |
| sp499 | 31.8068716 | -2.5689347 | 0.7101965 | -3.617217 | 0.0002978 | 0.0006785 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Rikenellaceae | RC9 gut group | 1054 |
| sp501 | 125.6584147 | -4.6197504 | 0.6234720 | -7.409715 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Bacteroidia | Bacteroidales | Rikenellaceae | Unclassified Rikenellaceae | 4289 |
| sp506 | 13.7786985 | -4.8154924 | 0.7516860 | -6.406256 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Class Incertae Sedis | Order Incertae Sedis | Family Incertae Sedis | Prolixibacter | 433 |
| sp516 | 18.7973547 | -6.6515369 | 0.9678025 | -6.872825 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Cytophagia | Cytophagales | Cyclobacteriaceae | Unclassified Cyclobacteriaceae | 566 |
| sp528 | 40.4263161 | -8.4599877 | 0.9412466 | -8.988067 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Cytophagia | Cytophagales | Cytophagaceae | Meniscus | 1283 |
| sp531 | 6.3339721 | -5.7778325 | 1.3468706 | -4.289820 | 0.0000179 | 0.0000575 | Bacteria | Bacteroidetes | Cytophagia | Cytophagales | Cytophagaceae | Pontibacter | 322 |
| sp536 | 11.7716589 | -7.0383217 | 1.1816636 | -5.956282 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Cytophagia | Cytophagales | Cytophagaceae | Sporocytophaga | 313 |
| sp537 | 23.3403766 | -4.2290026 | 0.7029461 | -6.016112 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Cytophagia | Cytophagales | Cytophagaceae | Unclassified Cytophagaceae | 581 |
| sp583 | 6.1113018 | -5.7420258 | 1.1004836 | -5.217730 | 0.0000002 | 0.0000007 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Cryomorphaceae | NS7 marine group | 209 |
| sp585 | 9.8300527 | -2.2521462 | 0.9637371 | -2.336889 | 0.0194450 | 0.0309923 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Cryomorphaceae | Unclassified Cryomorphaceae | 313 |
| sp595 | 23.4713522 | -5.9928971 | 0.8840498 | -6.778914 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Capnocytophaga | 769 |
| sp597 | 46.8710224 | 1.2725668 | 0.5043307 | 2.523278 | 0.0116266 | 0.0196506 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Chryseobacterium | 2013 |
| sp607 | 165.6320428 | 3.3389239 | 0.4225088 | 7.902613 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Flavobacterium | 8142 |
| sp629 | 11.3974097 | -5.2499455 | 0.8564924 | -6.129588 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Ornithobacterium | 344 |
| sp631 | 19.2564121 | -7.0339038 | 0.9274219 | -7.584362 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Polaribacter | 639 |
| sp644 | 282.6908383 | -0.9449575 | 0.3565166 | -2.650529 | 0.0080366 | 0.0139767 | Bacteria | Bacteroidetes | Flavobacteria | Flavobacteriales | Flavobacteriaceae | Unclassified Flavobacteriaceae | 10425 |
| sp677 | 22.2242683 | -4.5649976 | 0.6392086 | -7.141640 | 0.0000000 | 0.0000000 | Bacteria | Bacteroidetes | Sphingobacteriia | Sphingobacteriales | Sphingobacteriaceae | Mucilaginibacter | 697 |
| sp681 | 22.6285544 | -1.6371412 | 0.5917693 | -2.766519 | 0.0056658 | 0.0102497 | Bacteria | Bacteroidetes | Sphingobacteriia | Sphingobacteriales | Sphingobacteriaceae | Pedobacter | 739 |
| sp684 | 16.3224356 | -1.1034331 | 0.4749619 | -2.323203 | 0.0201682 | 0.0319849 | Bacteria | Bacteroidetes | Sphingobacteriia | Sphingobacteriales | Sphingobacteriaceae | Sphingobacterium | 601 |
| sp685 | 64.6554916 | -2.8609085 | 0.5112077 | -5.596372 | 0.0000000 | 0.0000001 | Bacteria | Bacteroidetes | Sphingobacteriia | Sphingobacteriales | Sphingobacteriaceae | Unclassified Sphingobacteriaceae | 2174 |
| sp715 | 1.1158274 | -2.6818752 | 1.0531095 | -2.546625 | 0.0108770 | 0.0185579 | Bacteria | Chlorobi | Chlorobia | Chlorobiales | Chlorobiaceae | Unclassified Chlorobiaceae | 117 |
| sp716 | 4.0478543 | -5.1743511 | 0.9619788 | -5.378862 | 0.0000001 | 0.0000003 | Bacteria | Chlorobi | Chlorobia | Chlorobiales | OPB56 | Chlorobi | 150 |
| sp726 | 3.4245305 | -4.5668720 | 1.2775150 | -3.574809 | 0.0003505 | 0.0007741 | Bacteria | Chloroflexi | Anaerolineae | Anaerolineales | Anaerolineaceae | Leptolinea | 130 |
| sp728 | 2.5692738 | -4.9085094 | 1.4040111 | -3.496062 | 0.0004722 | 0.0010240 | Bacteria | Chloroflexi | Anaerolineae | Anaerolineales | Anaerolineaceae | Longilinea | 116 |
| sp729 | 53.5090982 | -6.2975387 | 0.9025149 | -6.977767 | 0.0000000 | 0.0000000 | Bacteria | Chloroflexi | Anaerolineae | Anaerolineales | Anaerolineaceae | Unclassified Anaerolineaceae | 1817 |
| sp823 | 7.6258543 | -4.1697887 | 0.7283677 | -5.724841 | 0.0000000 | 0.0000000 | Bacteria | Cyanobacteria | Cyanobacteria | SubsectionIII | FamilyI | Unclassified FamilyI | 271 |
| sp870 | 4.5969889 | -3.5885230 | 0.9129094 | -3.930864 | 0.0000846 | 0.0002257 | Bacteria | Elusimicrobia | Elusimicrobia | Elusimicrobiales | Elusimicrobiaceae | Elusimicrobium | 142 |
| sp873 | 8.9486702 | -4.0455212 | 1.1873620 | -3.407151 | 0.0006564 | 0.0013740 | Bacteria | Fibrobacteres | Fibrobacteria | Fibrobacterales | Fibrobacteraceae | Fibrobacter | 320 |
| sp927 | 6.4447652 | -1.6363732 | 0.6982061 | -2.343682 | 0.0190944 | 0.0306875 | Bacteria | Firmicutes | Bacilli | Bacillales | Paenibacillaceae | Unclassified Paenibacillaceae | 202 |
| sp951 | 5.8593773 | -4.2025190 | 0.9113628 | -4.611247 | 0.0000040 | 0.0000136 | Bacteria | Firmicutes | Bacilli | Bacillales | Staphylococcaceae | Staphylococcus | 362 |
| sp985 | 117.7002799 | 7.5106437 | 0.7653132 | 9.813816 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Carnobacteriaceae | Trichococcus | 4784 |
| sp986 | 2.4732092 | 4.8578046 | 0.9204929 | 5.277395 | 0.0000001 | 0.0000005 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Carnobacteriaceae | Unclassified Carnobacteriaceae | 105 |
| sp987 | 295.7420589 | 6.5181385 | 0.5483404 | 11.887029 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Enterococcaceae | Enterococcus | 13622 |
| sp991 | 28.3998164 | 4.5575296 | 0.5802559 | 7.854344 | 0.0000000 | 0.0000000 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Enterococcaceae | Unclassified Enterococcaceae | 3361 |
| sp996 | 6.0683408 | -2.8749654 | 0.7792238 | -3.689525 | 0.0002247 | 0.0005374 | Bacteria | Firmicutes | Bacilli | Lactobacillales | Lactobacillaceae | Unclassified Lactobacillaceae | 213 |
# Phylum
x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Phylum, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaxa$Phylum = factor(as.character(sigtab_metaxa$Phylum), levels = names(x))
# Class
x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaxa$Class = factor(as.character(sigtab_metaxa$Class), levels = names(x))
# Genus
x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Genus, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaxa$Genus = factor(as.character(sigtab_metaxa$Genus), levels = names(x))
sorted_sigtab <- sigtab_metaxa[order(-sigtab_metaxa$log2FoldChange), ]
# Save Ben
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaxa2_DESeq2_Ben_Fin.txt",
# row.names=T, sep = "\t", col.names = T)
metaxa_Fin <- subset(sorted_sigtab,log2FoldChange>=0)
metaxa_Ben <- subset(sorted_sigtab,log2FoldChange<=0)
metaxa_Ben <- metaxa_Ben[order(metaxa_Ben$log2FoldChange), ]
head(metaxa_Fin)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 233 sp985 117.700280 7.510644 0.7653132 9.813816 9.818344e-23
## 2 sp1002 226.471157 7.147466 0.7521390 9.502852 2.042194e-21
## 235 sp987 295.742059 6.518138 0.5483404 11.887029 1.382369e-32
## 89 sp1666 7.175548 5.108515 0.9932698 5.143129 2.701994e-07
## 234 sp986 2.473209 4.857805 0.9204929 5.277395 1.310334e-07
## 236 sp991 28.399816 4.557530 0.5802559 7.854344 4.018686e-15
## padj Domain Phylum Class Order
## 233 8.836510e-21 Bacteria Firmicutes Bacilli Lactobacillales
## 2 1.225316e-19 Bacteria Firmicutes Bacilli Lactobacillales
## 235 4.976527e-30 Bacteria Firmicutes Bacilli Lactobacillales
## 89 1.013248e-06 Bacteria Proteobacteria Betaproteobacteria Burkholderiales
## 234 5.360458e-07 Bacteria Firmicutes Bacilli Lactobacillales
## 236 9.644845e-14 Bacteria Firmicutes Bacilli Lactobacillales
## Family Genus n
## 233 Carnobacteriaceae Trichococcus 4784
## 2 Streptococcaceae Lactococcus 6649
## 235 Enterococcaceae Enterococcus 13622
## 89 Comamonadaceae Sphaerotilus 476
## 234 Carnobacteriaceae Unclassified Carnobacteriaceae 105
## 236 Enterococcaceae Unclassified Enterococcaceae 3361head(metaxa_Ben)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 85 sp1648 48.09027 -8.704075 1.5683988 -5.549657 2.862311e-08
## 146 sp2016 276.42508 -8.631744 0.9915256 -8.705518 3.161283e-18
## 111 sp1798 53.72342 -8.504604 1.1338236 -7.500817 6.342129e-14
## 207 sp528 40.42632 -8.459988 0.9412466 -8.988067 2.516168e-19
## 118 sp1833 40.46754 -8.096920 1.4034686 -5.769221 7.963890e-09
## 133 sp1906 117.50842 -7.962389 1.1291554 -7.051632 1.768311e-12
## padj Domain Phylum Class
## 85 1.241484e-07 Bacteria Proteobacteria Betaproteobacteria
## 146 1.264513e-16 Bacteria Proteobacteria Gammaproteobacteria
## 111 1.098879e-12 Bacteria Proteobacteria Deltaproteobacteria
## 207 1.294029e-17 Bacteria Bacteroidetes Cytophagia
## 118 3.927398e-08 Bacteria Proteobacteria Deltaproteobacteria
## 133 1.929066e-11 Bacteria Proteobacteria Epsilonproteobacteria
## Order Family Genus n
## 85 Burkholderiales Comamonadaceae Macromonas 1495
## 146 Chromatiales Halothiobacillaceae Thiovirga 7600
## 111 Desulfobacterales Desulfobacteraceae Unclassified Desulfobacteraceae 1612
## 207 Cytophagales Cytophagaceae Meniscus 1283
## 118 Desulfuromonadales BVA18 Geobacter 1644
## 133 Campylobacterales Helicobacteraceae Sulfuricurvum 2740Plot
Benin-Finland
# Genus
#metaxa_Fin20 <- metaxa_Fin[1:20,]
#metaxa_Ben20 <- metaxa_Ben[1:20,]
#metaxa_top_sigtab <- rbind(metaxa_Fin20, metaxa_Ben20)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 15)
#p <- ggplot(data = metaxa_top_sigtab,aes(x = Genus, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) + ylab("Log2 Fold Change\nBenin Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Enriched taxa in Finnish / Beninise samples (20 taxa with highest log-2-fold change for each country)") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 14),
# axis.text.y = element_text(family = "Times", size = 16, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold"),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 20, face = "italic"),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 20, face = "bold"))
#p
#ggsave(filename = "FIN_BEN_deseq40_metaxa.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# ESKAPE etc
#metaxa_selected_sigtable <- subset(sigtab_metaxa, Genus =="Staphylococcus" | Genus == "Escherichia-Shigella" | Genus == "Enterococcus" | Genus == "Acinetobacter" | Genus == "Enterobacter" | Genus == "Pseudomonas" | Genus == "Clostridium" | Genus == "Citrobacter" | Genus == "Streptococcus" | Genus == "Aeromonas" | Genus == "Stenotrophomonas" | Genus == "Campylobacter" | Genus == "Corynebacterium" | Genus == "Mycobacterium" | Genus == "Neisseria" | Genus == "Salmonella" | Genus == "Vibrio" | Genus == "Klebsiella" | Genus =="Unclassified Staphylococcaceae" | Genus == "Unclassified Enterobacteriaceae" | Genus == "Unclassified Enterococcaceae" | Genus == "Unclassified Pseudomonadaceae" | Genus =="Unclassified Clostridiaceae"| Genus == "Unclassified Streptococcaceae" | Genus == "Unclassified Aeromonadaceae"| Genus == "Unclassified Campylobacteraceae" | Genus == "Unclassified Corynebacteriaceae" | Genus == "Unclassified Mycobacteriaceae" | Genus == "Unclassified Moraxellaceae" | Genus == "Unclassified Vibrionaceae"| Genus == "Unclassified Xanthomonadaceae" | Genus == "Unclassified Neisseriaceae")
#sorted_metaxa_selected_sigtable <- metaxa_selected_sigtable[order(-metaxa_selected_sigtable$log2FoldChange), ]
#selected_metaxa_Fin <- subset(sorted_metaxa_selected_sigtable,log2FoldChange>=0)
#selected_metaxa_Ben <- subset(sorted_metaxa_selected_sigtable,log2FoldChange<=0)
#selected_metaxa_Ben <- selected_metaxa_Ben[order(selected_metaxa_Ben$log2FoldChange), ]
#Fin11 <- selected_metaxa_Fin[1:11,]
#Ben9 <- selected_metaxa_Ben[1:9,]
#top_sigtab <- rbind(Fin11, Ben9)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 8)
#p <- ggplot(data = top_sigtab,aes(x = Genus, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) + ylab("Log2 Fold Change\nBenin Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Differentially abundant clinically relevant taxa in\n Finnish / Beninise hospital WW") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 14),
# axis.text.y = element_text(family = "Times", size = 24, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold"),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 26, face = "italic"),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 28, face = "bold"))
#p
#ggsave(filename = "FIN_BEN_deseq_selected_metaxa.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)DESeq2
Taxonomy (Metaxa2), Genus
#metaxa_deseq_PHY <- prune_taxa(taxa_sums(metaxa_PHY_stat) > 0, metaxa_PHY_stat)
#metaxa_deseq_PHY <- tax_glom(metaxa_deseq_PHY, taxrank = "Genus")
# Take pair wise comparisons
#deseq_PHY = subset_samples(metaxa_deseq_PHY, country == "Burkina Faso" | country == "Finland")
#varianceThreshold = 50
#keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
#deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
#deseq_PHY
#dds = phyloseq_to_deseq2(deseq_PHY, ~country)
#ds$category <- relevel(dds$country, "Burkina Faso", "Finland")
#dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)
#res = results(dds, cooksCutoff = FALSE, alpha = 0.05)
#resultsNames(dds)
#plotDispEsts(dds)
#head(res)
#summary(res)
#res = res[order(res$padj, na.last=NA), ]
#alpha = 0.05
#sigtab_metaxa = res[which(res$padj < alpha), ]
#sigtab_metaxa = cbind(as(sigtab_metaxa, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaxa), ], "matrix"))
#n <- rowSums(otu_table(deseq_PHY))
#sigtab_metaxa = merge(sigtab_metaxa, as.data.frame(n), by = 0)
#kable(sigtab_metaxa, caption = "Taxonomy")
# Phylum
#x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Phylum, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_metaxa$Phylum = factor(as.character(sigtab_metaxa$Phylum), levels = names(x))
# Class
#x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Class, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_metaxa$Class = factor(as.character(sigtab_metaxa$Class), levels = names(x))
# Genus
#x = tapply(sigtab_metaxa$log2FoldChange, sigtab_metaxa$Genus, function(x) max(x))
#x = sort(x, TRUE)
#sigtab_metaxa$Genus = factor(as.character(sigtab_metaxa$Genus), levels = names(x))
#sorted_sigtab <- sigtab_metaxa[order(-sigtab_metaxa$log2FoldChange), ]
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaxa2_DESeq2_BF_Fin.txt",
# row.names=T, sep = "\t", col.names = T)Plot
BF-Finland
#metaxa_Fin <- subset(sorted_sigtab,log2FoldChange>=0)
#metaxa_BF <- subset(sorted_sigtab,log2FoldChange<=0)
#metaxa_BF <- metaxa_BF[order(metaxa_BF$log2FoldChange), ]
# Genus
#metaxa_Fin20 <- metaxa_Fin[1:20,]
#metaxa_BF20 <- metaxa_BF[1:20,]
#metaxa_top_sigtab <- rbind(metaxa_Fin20, metaxa_BF20)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 15)
#p <- ggplot(data = metaxa_top_sigtab,aes(x = Genus, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) + ylab("Log2 Fold Change\nBurkina Faso Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Enriched taxa in Finnish / Burkinabe samples (20 taxa with highest log-2-fold change for each country)") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 14),
# axis.text.y = element_text(family = "Times", size = 16, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold"),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 20, face = "italic"),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 20, face = "bold"))
#p
#ggsave(filename = "FIN_BF_deseq40_metaxa.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# ESKAPE etc
#metaxa_selected_sigtable <- subset(sigtab_metaxa, Genus =="Staphylococcus" | Genus == "Escherichia-Shigella" | Genus == "Enterococcus" | Genus == "Acinetobacter" | Genus == "Enterobacter" | Genus == "Pseudomonas" | Genus == "Clostridium" | Genus == "Citrobacter" | Genus == "Streptococcus" | Genus == "Aeromonas" | Genus == "Stenotrophomonas" | Genus == "Campylobacter" | Genus == "Corynebacterium" | Genus == "Mycobacterium" | Genus == "Neisseria" | Genus == "Salmonella" | Genus == "Vibrio" | Genus == "Klebsiella" | Genus =="Unclassified Staphylococcaceae" | Genus == "Unclassified Enterobacteriaceae" | Genus == "Unclassified Enterococcaceae" | Genus == "Unclassified Pseudomonadaceae" | Genus =="Unclassified Clostridiaceae"| Genus == "Unclassified Streptococcaceae" | Genus == "Unclassified Aeromonadaceae"| Genus == "Unclassified Campylobacteraceae" | Genus == "Unclassified Corynebacteriaceae" | Genus == "Unclassified Mycobacteriaceae" | Genus == "Unclassified Moraxellaceae" | Genus == "Unclassified Vibrionaceae"| Genus == "Unclassified Xanthomonadaceae" | Genus == "Unclassified Neisseriaceae")
#sorted_metaxa_selected_sigtable <- metaxa_selected_sigtable[order(-metaxa_selected_sigtable$log2FoldChange), ]
#metaxa_Fin <- subset(sorted_metaxa_selected_sigtable,log2FoldChange>=0)
#metaxa_BF <- subset(sorted_metaxa_selected_sigtable,log2FoldChange<=0)
#metaxa_BF <- metaxa_BF[order(metaxa_BF$log2FoldChange), ]
#Fin5 <- metaxa_Fin[1:5,]
#BF9 <- metaxa_BF[1:9,]
#top_sigtab <- rbind(Fin5, BF9)
#cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 8)
#p <- ggplot(data = top_sigtab,aes(x = Genus, y = log2FoldChange)) +
# geom_bar(stat = "identity",
# aes(fill = Class), position = position_dodge(preserve = "single")) + ylab("Log2 Fold Change\nBurkina Faso Finland") +
# scale_fill_manual(values = cols) +
# ggtitle("Differentially abundant clinically relevant taxa in \nFinnish / Burkinabe hospital WW") +
# coord_flip() +
# theme_minimal() +
# theme(axis.text.x = element_text(family = "Times", size = 14),
# axis.text.y = element_text(family = "Times", size = 24, face = "bold.italic"),
# axis.title.x = element_text(family = "Times", size = 20, face = "bold"),
# axis.title.y = element_blank(),
# legend.text = element_text(family = "Times", size = 26, face = "italic"),
# legend.title = element_blank(),
# plot.title = element_text(family = "Times", size = 28, face = "bold"))
#p
#ggsave(filename = "FIN_BF_deseq_selected_metaxa.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)DESeq2
Taxonomy (Metaphlan3), Genus/Species
Benin-Finland
# Load data
OTU_metaphlan <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/mod_merged_abundance_table_species.txt", header=T)
# Make sure tax tabe is in order
#tax_table_metaphlan <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", quote="\"", comment.char="")
#identical(tax_table_metaphlan$V1, OTU_metaphlan$clade_name)
tax_table_metaphlan <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", header=FALSE, sep=";")
colnames(tax_table_metaphlan) <- c("Kingdom", "Phylum", "Class", "Order", "Family", "Genus", "Species")
# Remove "__"
tax_table_metaphlan <- apply(tax_table_metaphlan, 2, function(y) (gsub(".__", "", y)))
match <- match(rownames(metadata), colnames(OTU_metaphlan))
OTU_metaphlan <- OTU_metaphlan[,match]
all(rownames(metadata) == colnames(OTU_metaphlan))## [1] TRUEOTU_metaphlan_deseq = OTU_metaphlan
vec <- as.vector(metadata$SSU_counts)
deseq_OTU <- mapply(FUN = `*`, as.data.frame(OTU_metaphlan_deseq), vec)
metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(tax_table_metaphlan)))
# OLD WAY
#deseq_OTU <- OTU_metaphlan[, ] * 10^5 + 1
#metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
# tax_table(as.matrix(tax_table_metaphlan)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaphlan_deseq = subset_samples(metaphlan_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaphlan_deseq <- subset_samples(metaphlan_deseq, country != "UK")
## Exclude biological / technical replicates
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
metaphlan_deseq_stat <- prune_taxa(taxa_sums(metaphlan_deseq_stat) > 0, metaphlan_deseq_stat)
# Take pair wise comparisons
deseq_PHY = subset_samples(metaphlan_deseq_stat, country == "Benin" | country == "Finland")
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 677 taxa and 33 samples ]
## sample_data() Sample Data: [ 33 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 677 taxa by 7 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer modedds$category <- relevel(dds$country, "Benin", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## -- replacing outliers and refitting for 537 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## fitting model and testingres = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Benin"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_metaphlan = res[which(res$padj < alpha), ]
sigtab_metaphlan = cbind(as(sigtab_metaphlan, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaphlan),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_metaphlan = merge(sigtab_metaphlan, as.data.frame(n), by = 0)
#kable(sigtab_metaphlan, caption = "Taxonomy")
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Class = factor(as.character(sigtab_metaphlan$Class), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Phylum, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Phylum = factor(as.character(sigtab_metaphlan$Phylum), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Genus, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Genus = factor(as.character(sigtab_metaphlan$Genus), levels = names(x))Save (Species)
Benin-Finland
sorted_sigtab <- sigtab_metaphlan[order(-sigtab_metaphlan$log2FoldChange), ]
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaphlan3_DESeq2_Ben_Fin_s.txt",
# row.names=T, sep = "\t", col.names = T)
head(sorted_sigtab)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 48 sp1237 1082.7292 30.00000 2.801522 10.708466 9.288845e-27
## 50 sp1245 569.9077 30.00000 3.391066 8.846776 9.008396e-19
## 149 sp511 1860.6666 30.00000 3.391061 8.846788 9.007418e-19
## 152 sp516 438.1770 30.00000 3.391068 8.846771 9.008820e-19
## 38 sp1179 269.3041 29.51292 3.391073 8.703119 3.228843e-18
## 93 sp225 391.1714 29.37676 3.021778 9.721681 2.437240e-22
## padj Kingdom Phylum Class Order
## 48 2.949147e-25 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## 50 1.680145e-17 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## 149 1.680145e-17 Bacteria Firmicutes Bacilli Lactobacillales
## 152 1.680145e-17 Bacteria Firmicutes Bacilli Lactobacillales
## 38 2.676352e-17 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## 93 6.993005e-21 Bacteria Bacteroidetes Bacteroidia Bacteroidales
## Family Genus Species n
## 48 Moraxellaceae Acinetobacter Acinetobacter_guillouiae 6
## 50 Moraxellaceae Acinetobacter Acinetobacter_lwoffii 4
## 149 Streptococcaceae Lactococcus Lactococcus_chungangensis 3
## 152 Streptococcaceae Lactococcus Lactococcus_raffinolactis 4
## 38 Enterobacteriaceae Raoultella Raoultella_ornithinolytica 5
## 93 Bacteroidaceae Bacteroides Bacteroides_faecis 5DESeq2
Taxonomy (Metaphlan3), Genus/Species
Benin-Finland
# Load data
OTU_metaphlan <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/mod_merged_abundance_table_species.txt", header=T)
# Make sure tax tabe is in order
#tax_table_metaphlan <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", quote="\"", comment.char="")
#identical(tax_table_metaphlan$V1, OTU_metaphlan$clade_name)
tax_table_metaphlan <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", header=FALSE, sep=";")
colnames(tax_table_metaphlan) <- c("Kingdom", "Phylum", "Class", "Order", "Family", "Genus", "Species")
# Remove "__"
tax_table_metaphlan <- apply(tax_table_metaphlan, 2, function(y) (gsub(".__", "", y)))
match <- match(rownames(metadata), colnames(OTU_metaphlan))
OTU_metaphlan <- OTU_metaphlan[,match]
all(rownames(metadata) == colnames(OTU_metaphlan))## [1] TRUEOTU_metaphlan_deseq = OTU_metaphlan
vec <- as.vector(metadata$SSU_counts)
deseq_OTU <- mapply(FUN = `*`, as.data.frame(OTU_metaphlan_deseq), vec)
metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(tax_table_metaphlan)))
# OLD WAY
#deseq_OTU <- OTU_metaphlan[, ] * 10^5 + 1
#metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
# tax_table(as.matrix(tax_table_metaphlan)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaphlan_deseq = subset_samples(metaphlan_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaphlan_deseq <- subset_samples(metaphlan_deseq, country != "UK")
## Exclude biological / technical replicates
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
metaphlan_deseq_stat <- prune_taxa(taxa_sums(metaphlan_deseq_stat) > 0, metaphlan_deseq_stat)
# Take pair wise comparisons
deseq_PHY = subset_samples(metaphlan_deseq_stat, country == "Benin" | country == "Finland")
# Get genus
deseq_PHY <- tax_glom(deseq_PHY, taxrank = "Genus")
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 284 taxa and 33 samples ]
## sample_data() Sample Data: [ 33 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 284 taxa by 7 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer modedds$category <- relevel(dds$country, "Benin", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## -- replacing outliers and refitting for 212 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## fitting model and testingres = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Benin"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_metaphlan = res[which(res$padj < alpha), ]
sigtab_metaphlan = cbind(as(sigtab_metaphlan, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaphlan),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_metaphlan = merge(sigtab_metaphlan, as.data.frame(n), by = 0)
#kable(sigtab_metaphlan, caption = "Taxonomy")
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Class = factor(as.character(sigtab_metaphlan$Class), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Phylum, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Phylum = factor(as.character(sigtab_metaphlan$Phylum), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Genus, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Genus = factor(as.character(sigtab_metaphlan$Genus), levels = names(x))Save and plot (Genus)
Fin_Benin
sorted_metaphlan_sigtable <- sigtab_metaphlan[order(-sigtab_metaphlan$log2FoldChange), ]
#write.table(sorted_metaphlan_sigtable, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaphlan3_DESeq2_Ben_Fin_g.txt",
# row.names=T, sep = "\t", col.names = T)
head(sorted_metaphlan_sigtable)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 25 sp1200 210.33648 28.78040 3.391092 8.487059 2.119320e-17
## 88 sp960 185.38452 28.60806 3.385660 8.449773 2.918688e-17
## 20 sp1173 81.11827 27.45000 3.391145 8.094612 5.744739e-16
## 49 sp258 76.64687 27.36962 3.391150 8.070896 6.978429e-16
## 78 sp779 73.85810 27.31714 3.391154 8.055411 7.921187e-16
## 21 sp1179 65.92663 27.15971 3.316690 8.188800 2.638430e-16
## padj Kingdom Phylum Class Order
## 25 2.007777e-16 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## 88 2.501732e-16 Bacteria Proteobacteria Betaproteobacteria Burkholderiales
## 20 4.136212e-15 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## 49 4.652286e-15 Bacteria Bacteroidetes Bacteroidia Bacteroidales
## 78 5.092192e-15 Bacteria Proteobacteria Alphaproteobacteria Caulobacterales
## 21 1.978822e-15 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## Family Genus Species n
## 25 Yersiniaceae Serratia <NA> 4
## 88 Burkholderiales_unclassified Sphaerotilus <NA> 4
## 20 Enterobacteriaceae Kluyvera <NA> 3
## 49 Barnesiellaceae Coprobacter <NA> 3
## 78 Caulobacteraceae Brevundimonas <NA> 4
## 21 Enterobacteriaceae Raoultella <NA> 6DESeq2
Taxonomy (Metaphlan3), Genus/Species
BF-Finland
# Load data
OTU_metaphlan <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/mod_merged_abundance_table_species.txt", header=T)
# Make sure tax tabe is in order
#tax_table_metaphlan <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", quote="\"", comment.char="")
#identical(tax_table_metaphlan$V1, OTU_metaphlan$clade_name)
tax_table_metaphlan <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", header=FALSE, sep=";")
colnames(tax_table_metaphlan) <- c("Kingdom", "Phylum", "Class", "Order", "Family", "Genus", "Species")
# Remove "__"
tax_table_metaphlan <- apply(tax_table_metaphlan, 2, function(y) (gsub(".__", "", y)))
match <- match(rownames(metadata), colnames(OTU_metaphlan))
OTU_metaphlan <- OTU_metaphlan[,match]
all(rownames(metadata) == colnames(OTU_metaphlan))## [1] TRUEOTU_metaphlan_deseq = OTU_metaphlan
vec <- as.vector(metadata$SSU_counts)
deseq_OTU <- mapply(FUN = `*`, as.data.frame(OTU_metaphlan_deseq), vec)
metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(tax_table_metaphlan)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaphlan_deseq = subset_samples(metaphlan_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaphlan_deseq <- subset_samples(metaphlan_deseq, country != "UK")
## Exclude biological / technical replicates
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
metaphlan_deseq_stat <- prune_taxa(taxa_sums(metaphlan_deseq_stat) > 0, metaphlan_deseq_stat)
# Take pair wise comparisons
deseq_PHY = subset_samples(metaphlan_deseq_stat, country == "Burkina Faso" | country == "Finland")
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 818 taxa and 42 samples ]
## sample_data() Sample Data: [ 42 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 818 taxa by 7 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer mode## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]dds$category <- relevel(dds$country, "Burkina Faso", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors
## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## final dispersion estimates## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## -- replacing outliers and refitting for 711 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]res = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Burkina.Faso"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_metaphlan = res[which(res$padj < alpha), ]
sigtab_metaphlan = cbind(as(sigtab_metaphlan, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaphlan),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_metaphlan = merge(sigtab_metaphlan, as.data.frame(n), by = 0)
#kable(sigtab_metaphlan, caption = "Taxonomy")
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Class = factor(as.character(sigtab_metaphlan$Class), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Phylum, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Phylum = factor(as.character(sigtab_metaphlan$Phylum), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Genus, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Genus = factor(as.character(sigtab_metaphlan$Genus), levels = names(x))Save (Species)
BF-Finland
sorted_sigtab <- sigtab_metaphlan[order(-sigtab_metaphlan$log2FoldChange), ]
#write.table(sorted_sigtab, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaphlan3_DESeq2_BF_Fin_s.txt",
# row.names=T, sep = "\t", col.names = T)
head(sorted_sigtab)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 31 sp1110 268.4197 30 3.690596 8.128769 4.336735e-16
## 47 sp1179 447.1906 30 3.690587 8.128788 4.336052e-16
## 58 sp1229 527.0725 30 3.690585 8.128792 4.335896e-16
## 63 sp1237 1797.9167 30 3.020567 9.931911 3.024042e-23
## 69 sp1245 946.3554 30 3.690580 8.128803 4.335510e-16
## 82 sp1295 269.4503 30 3.690596 8.128769 4.336729e-16
## padj Kingdom Phylum Class Order
## 31 6.709394e-15 Bacteria Proteobacteria Gammaproteobacteria Aeromonadales
## 47 6.709394e-15 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## 58 6.709394e-15 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## 63 1.247417e-21 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## 69 6.709394e-15 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## 82 6.709394e-15 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales
## Family Genus Species n
## 31 Aeromonadaceae Aeromonas Aeromonas_salmonicida 3
## 47 Enterobacteriaceae Raoultella Raoultella_ornithinolytica 5
## 58 Moraxellaceae Acinetobacter Acinetobacter_bereziniae 4
## 63 Moraxellaceae Acinetobacter Acinetobacter_guillouiae 6
## 69 Moraxellaceae Acinetobacter Acinetobacter_lwoffii 4
## 82 Pseudomonadaceae Pseudomonas Pseudomonas_fluorescens_group 3DESeq2
Taxonomy (Metaphlan3), Genus/Species
BF-Finland
# Load data
OTU_metaphlan <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/mod_merged_abundance_table_species.txt", header=T)
# Make sure tax tabe is in order
#tax_table_metaphlan <- read.table("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", quote="\"", comment.char="")
#identical(tax_table_metaphlan$V1, OTU_metaphlan$clade_name)
tax_table_metaphlan <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/tax_table_metaphlan", header=FALSE, sep=";")
colnames(tax_table_metaphlan) <- c("Kingdom", "Phylum", "Class", "Order", "Family", "Genus", "Species")
# Remove "__"
tax_table_metaphlan <- apply(tax_table_metaphlan, 2, function(y) (gsub(".__", "", y)))
match <- match(rownames(metadata), colnames(OTU_metaphlan))
OTU_metaphlan <- OTU_metaphlan[,match]
all(rownames(metadata) == colnames(OTU_metaphlan))## [1] TRUEOTU_metaphlan_deseq = OTU_metaphlan
vec <- as.vector(metadata$SSU_counts)
deseq_OTU <- mapply(FUN = `*`, as.data.frame(OTU_metaphlan_deseq), vec)
metaphlan_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(tax_table_metaphlan)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
metaphlan_deseq = subset_samples(metaphlan_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
metaphlan_deseq <- subset_samples(metaphlan_deseq, country != "UK")
## Exclude biological / technical replicates
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
metaphlan_deseq_stat <- subset_samples(metaphlan_deseq_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
metaphlan_deseq_stat <- prune_taxa(taxa_sums(metaphlan_deseq_stat) > 0, metaphlan_deseq_stat)
# Take pair wise comparisons
deseq_PHY = subset_samples(metaphlan_deseq_stat, country == "Burkina Faso" | country == "Finland")
# Get genus
deseq_PHY <- tax_glom(deseq_PHY, taxrank = "Genus")
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 332 taxa and 42 samples ]
## sample_data() Sample Data: [ 42 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 332 taxa by 7 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer mode## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]dds$category <- relevel(dds$country, "Burkina Faso", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors
## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## final dispersion estimates## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## -- replacing outliers and refitting for 285 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]## fitting model and testing## Note: levels of factors in the design contain characters other than
## letters, numbers, '_' and '.'. It is recommended (but not required) to use
## only letters, numbers, and delimiters '_' or '.', as these are safe characters
## for column names in R. [This is a message, not an warning or error]res = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Burkina.Faso"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_metaphlan = res[which(res$padj < alpha), ]
sigtab_metaphlan = cbind(as(sigtab_metaphlan, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_metaphlan),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_metaphlan = merge(sigtab_metaphlan, as.data.frame(n), by = 0)
#kable(sigtab_metaphlan, caption = "Taxonomy")
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Class = factor(as.character(sigtab_metaphlan$Class), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Phylum, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Phylum = factor(as.character(sigtab_metaphlan$Phylum), levels = names(x))
x = tapply(sigtab_metaphlan$log2FoldChange, sigtab_metaphlan$Genus, function(x) max(x))
x = sort(x, TRUE)
sigtab_metaphlan$Genus = factor(as.character(sigtab_metaphlan$Genus), levels = names(x))Save and plot (Genus)
Fin_BF
sorted_metaphlan_sigtable <- sigtab_metaphlan[order(-sigtab_metaphlan$log2FoldChange), ]
#write.table(sorted_metaphlan_sigtable, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/metaphlan3_Genus_DESeq2_BF_Fin_g.txt",
# row.names=T, sep = "\t", col.names = T)
head(sorted_metaphlan_sigtable)## Row.names baseMean log2FoldChange lfcSE stat pvalue
## 23 sp1179 895.36387 30.00000 3.256619 9.212009 3.200781e-20
## 42 sp187 155.47478 29.11599 3.690607 7.889215 3.040936e-15
## 3 sp1004 80.67191 28.12205 3.162094 8.893489 5.921947e-19
## 2 sp1003 49.07225 27.49675 3.619882 7.596033 3.053472e-14
## 41 sp179 31.80415 26.89334 3.690736 7.286717 3.176000e-13
## 76 sp755 24.67132 26.53262 3.690782 7.188889 6.532059e-13
## padj Kingdom Phylum Class Order
## 23 8.762139e-19 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales
## 42 3.027114e-14 Bacteria Actinobacteria Actinobacteria Propionibacteriales
## 3 1.296906e-17 Bacteria Proteobacteria Betaproteobacteria Burkholderiales
## 2 2.157130e-13 Bacteria Proteobacteria Betaproteobacteria Burkholderiales
## 41 1.987268e-12 Bacteria Actinobacteria Actinobacteria Nakamurellales
## 76 3.778950e-12 Bacteria Firmicutes Tissierellia Tissierellales
## Family Genus Species n
## 23 Enterobacteriaceae Raoultella <NA> 6
## 42 Propionibacteriaceae Propionibacterium <NA> 3
## 3 Comamonadaceae Ottowia <NA> 5
## 2 Comamonadaceae Melaminivora <NA> 4
## 41 Nakamurellaceae Nakamurella <NA> 3
## 76 Tissierellaceae Tissierella <NA> 4DESeq2
MGE (Mobilome)
Benin-Finland / BF-Finland
# Deseq
deseq_OTU <- OTU_MGE_length_SSU_norm[, ] * 10^5 + 1
MGE_deseq <- phyloseq(otu_table(deseq_OTU, taxa_are_rows = T), sample_data(metadata),
tax_table(as.matrix(MGE_tax_table_trim)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
MGE_deseq = subset_samples(MGE_deseq, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Exclude samples from UK hospital whose sample collection does not match with other samples
MGE_deseq <- subset_samples(MGE_deseq, country != "UK")
## Exclude biological / technical replicates
MGE_deseq_stat <- subset_samples(MGE_deseq, alias != "BH31" & alias != "BH33" & alias != "BH34B" & alias != "BH10"
& alias != "BFH38B" & alias != "FH8" & alias != "BH45" & alias != "BH59" & alias != "BH62")
# Create phyloseq object with only hospital WW samples sequenced here
MGE_deseq_stat <- subset_samples(MGE_deseq_stat, category == "WA hospital effluent" | category == "North Eu hospital effluent")
MGE_deseq_stat <- prune_taxa(taxa_sums(MGE_deseq_stat) > 0, MGE_deseq_stat)
# Take pair wise comparisons
deseq_PHY = subset_samples(MGE_deseq_stat, country == "Benin" | country == "Finland")
#deseq_PHY = subset_samples(MGE_deseq_stat, country == "Burkina Faso" | country == "Finland")
varianceThreshold = 50
keepOTUs = apply(otu_table(deseq_PHY), 1, var) > varianceThreshold
deseq_PHY = prune_taxa(keepOTUs, deseq_PHY)
deseq_PHY## phyloseq-class experiment-level object
## otu_table() OTU Table: [ 1122 taxa and 33 samples ]
## sample_data() Sample Data: [ 33 samples by 35 sample variables ]
## tax_table() Taxonomy Table: [ 1122 taxa by 3 taxonomic ranks ]dds = phyloseq_to_deseq2(deseq_PHY, ~country)## converting counts to integer modedds$category <- relevel(dds$country, "Benin", "Finland")
#dds$category <- relevel(dds$country, "Burkina Faso", "Finland")
dds = DESeq(dds, fitType = "mean", test = "Wald", betaPrior = FALSE)## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## -- replacing outliers and refitting for 394 genes
## -- DESeq argument 'minReplicatesForReplace' = 7
## -- original counts are preserved in counts(dds)## estimating dispersions## fitting model and testingres = results(dds, cooksCutoff = FALSE, alpha = 0.05)
resultsNames(dds)## [1] "Intercept" "country_Finland_vs_Benin"#plotDispEsts(dds)
#head(res)
#summary(res)
res = res[order(res$padj, na.last=NA), ]
alpha = 0.05
sigtab_MGE = res[which(res$padj < alpha), ]
sigtab_MGE = cbind(as(sigtab_MGE, "data.frame"), as(tax_table(deseq_PHY)[rownames(sigtab_MGE),
], "matrix"))
otu_table(deseq_PHY)[otu_table(deseq_PHY) == 1] <- 0
otu_table(deseq_PHY)[otu_table(deseq_PHY) > 0] <- 1
n <- rowSums(otu_table(deseq_PHY))
sigtab_MGE = merge(sigtab_MGE, as.data.frame(n), by = 0)
#kable(sigtab_MGE, caption = "Taxonomy")
x = tapply(sigtab_MGE$log2FoldChange, sigtab_MGE$Class, function(x) max(x))
x = sort(x, TRUE)
sigtab_MGE$Class = factor(as.character(sigtab_MGE$Class), levels = names(x))
x = tapply(sigtab_MGE$log2FoldChange, sigtab_MGE$Element, function(x) max(x))
x = sort(x, TRUE)
sigtab_MGE$Element = factor(as.character(sigtab_MGE$Element), levels = names(x))
# Sort
sorted_sigtab <- sigtab_MGE[order(-sigtab_MGE$log2FoldChange), ]Heatmap, ResFinder
## Load data as above
OTU_resfinder <-as.matrix(read.table("ARG_genemat.txt", header= T, check.names = F, row.names = 1))
# Reorder to match metadata
match <- match(rownames(metadata), colnames(OTU_resfinder))
OTU_resfinder <- OTU_resfinder[,match]
all(colnames(OTU_resfinder) == rownames(metadata))## [1] TRUE# Tax_table (Cluster names created using cd-hit and 90 % identity and added manually to the tax table in excel)
clusters_tax_table_resfinder <- read.csv("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table.txt", header=FALSE, sep=";")
colnames(clusters_tax_table_resfinder) <- c("Gene", "Cluster_name", "Class")
# Reorder columns
col_order <- c("Class", "Cluster_name", "Gene")
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[, col_order]
# Reorder tax_table to match OTU_resfinder
match <- match(rownames(OTU_resfinder), clusters_tax_table_resfinder$Gene)
clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
all(rownames(OTU_resfinder) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Divide by ARG gene lengths
## Get the lengths in terminal
# seqkit fx2tab --length --name --header-line resfinder.fasta > resfinder_lengths.txt
resfinder_lengths <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/resfinder_lengths.txt", header=FALSE, comment.char="#")
all(rownames(clusters_tax_table_resfinder$Gene) == resfinder_lengths$V1)## [1] TRUEOTU_resfinder_length_norm <- OTU_resfinder/resfinder_lengths[, 2]
# Normalization with Metaxa2 SSU counts
OTU_resfinder_length_SSU_norm <- t(t(OTU_resfinder_length_norm)/metadata$SSU_counts) * 1540
all(rownames(metadata) == colnames(OTU_resfinder_length_SSU_norm))## [1] TRUEidentical(OTU_resfinder_length_norm[2025, 5]/metadata$SSU_counts[5], OTU_resfinder_length_SSU_norm[2025, 5])## [1] TRUEall(rownames(OTU_resfinder_length_norm) == clusters_tax_table_resfinder$Gene)## [1] TRUE# Merge ARGs so that cluster names instead of gene names can be used
clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/clusters_tax_table_resfinder.txt")
all(clusters_tax_table_resfinder$Gene == rownames(OTU_resfinder_length_SSU_norm))## [1] TRUEClass <- as.data.frame(clusters_tax_table_resfinder$Class)
Cluster_name <- as.data.frame(clusters_tax_table_resfinder$Cluster_name)
Gene <- as.data.frame(clusters_tax_table_resfinder$Gene)
heat_length_SSU_norm <- cbind(OTU_resfinder_length_SSU_norm, Class, Cluster_name, Gene)
names(heat_length_SSU_norm)[names(heat_length_SSU_norm) == "clusters_tax_table_resfinder$Class"] <- "Class"
names(heat_length_SSU_norm)[names(heat_length_SSU_norm) == "clusters_tax_table_resfinder$Cluster_name"] <- "Cluster_name"
names(heat_length_SSU_norm)[names(heat_length_SSU_norm) == "clusters_tax_table_resfinder$Gene"] <- "Gene"
all(heat_length_SSU_norm$Gene == clusters_tax_table_resfinder$Gene)## [1] TRUE# Merge
merged <- ddply(heat_length_SSU_norm,"Cluster_name",numcolwise(sum))
# Remove duplicates
uniq_clust <- clusters_tax_table_resfinder[!duplicated(clusters_tax_table_resfinder[ , c("Cluster_name")]),]
# Reorder to match OTU
match <- match(uniq_clust$Cluster_name, merged$Cluster_name)
merged <- merged[match,]
all(merged$Cluster_name == uniq_clust$Cluster_name)## [1] TRUE# Tax table
match <- match(merged$Cluster_name, clusters_tax_table_resfinder$Cluster_name)
heat_clusters_tax_table_resfinder <- clusters_tax_table_resfinder[match,]
identical(merged$Cluster_name, heat_clusters_tax_table_resfinder$Cluster_name)## [1] TRUE# Reload to get new row numbers
write.table(merged, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_OTU_resfinder.txt",
row.names=T, sep = "\t", col.names = T)
heat_OTU_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_OTU_resfinder.txt", row.names=NULL)
heat_OTU_resfinder$row.names<-NULL
heat_OTU_resfinder$Cluster_name <- NULL
# Reload to get new row numbers
write.table(heat_clusters_tax_table_resfinder, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_clusters_tax_table_resfinder.txt", row.names=F, sep = "\t", col.names = T)
heat_clusters_tax_table_resfinder <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_clusters_tax_table_resfinder.txt", row.names=NULL)
# Create phyloseq
heat_resfinder_PHY <- phyloseq(otu_table(heat_OTU_resfinder, taxa_are_rows = TRUE), sample_data(metadata),
tax_table(as.matrix(heat_clusters_tax_table_resfinder)))
# Create phyloseq object with all samples
# Exclude samples with low sequencing quality (BFH24_S142, BH63_S118, FH10_S171)
heat_resfinder_PHY = subset_samples(heat_resfinder_PHY, alias != "BFH24" & alias != "BH63" & alias != "FH10")
# Subset samples
heat_resfinder_PHY <- subset_samples(heat_resfinder_PHY, category == "WA hospital effluent" |
category == "WA treated, receiving municipality channel" |
category == "WA treated, receiving river" |
category == "WA empty hospital septic tank" | category == "WA treated, drinking water" |
category == "WA street gutter" | category == "WA street gutter, sediment" |
category == "WA in-patient feces" | category == "WA treated, hand washing" |
category == "WA treated, hand washing" | category == "WA river, drinking water" |
category == "North Eu hospital effluent")Plot Heatmap
# Exctract ARGs
heat_resfinder_PHY_Clusters <- tax_glom(heat_resfinder_PHY, taxrank = "Cluster_name")
selected <- subset_taxa(heat_resfinder_PHY_Clusters,
Cluster_name == "blaOXA-370_1_clust" | Cluster_name == "blaIMP-58_1_clust" |
Cluster_name == "blaNDM-18_1_clust" | Cluster_name == "blaCTX-M-211_1_clust" |
Cluster_name == "blaCTX-M-4_1_clust" | Cluster_name == "blaCTX-M-110_1_clust" |
Cluster_name == "blaKPC-34_1_clust" | Cluster_name == "catB3_2_clust" |
Cluster_name == "blaGES-1_1_clust" | Cluster_name == "blaVEB-1_1_clust" |
Cluster_name == "cfxA_1_clust" | Cluster_name == "blaCMY-150_2_clust" |
Cluster_name == "aph(6)-Id_1_clust" | Cluster_name == "aadA2_1_clust" |
Cluster_name == "cmlA1_1_clust" | Cluster_name == "mcr-3.1_1_clust" |
Cluster_name == "mph(E)_1_clust" | Cluster_name == "erm(B)_18_clust" |
Cluster_name == "mcr-1.11_1_clust" | Cluster_name == "qnrA1_1_clust" |
Cluster_name == "nimA_1" | Cluster_name == "nimE_1_clust" |
Cluster_name == "sul1_11_clust" | Cluster_name == "mcr-5.1_1_clust" |
Cluster_name == "tet(A)_6_clust" |
Cluster_name == "dfrA1_1_clust" | Cluster_name == "erm(33)_1" |
Cluster_name == "mph(E)_1_clust" | Cluster_name == "blaOXA-368_1_clust" |
Cluster_name == "blaVIM-48_1_clust" | Cluster_name == "blaOXA-392_1_clust")
# OTU matrix
heat_OTU = as(otu_table(selected), "matrix")
# Coerce to data.frame
heat.df = as.data.frame(heat_OTU)
# Tax table matrix
heat_tax = as(tax_table(selected), "matrix")
# Save to get it back without the current first header row
write.table(heat_tax, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_tax", row.names=T, sep = "\t", col.names = F)
heat_tax <- read.delim("~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/heat_tax", row.names=1, header = F)
# Into dataframe
heat_tax.df = as.data.frame(heat_tax)
# Swap colnames
match <- match(rownames(heat.df), rownames(heat_tax.df))
temp <- heat_tax.df[match,]
rownames(heat.df) <- temp$V3
#metadata_row <- data.frame(temp$V2, row.names=rownames(heat.df))
# Col annotation
category <- as.matrix(sample_data(selected)[["category"]])
category <- as.factor(category)
category <- data.frame(category)
colnames(category) <- c("category")
rownames(category) <- as.matrix(colnames(otu_table(selected)))
ann_colors <- list(category = c("WA hospital effluent" = "#F04744",
"WA treated, receiving municipality channel" = "#AF32D9",
"WA treated, receiving river" = "black",
"WA empty hospital septic tank" = "#B1B6FC",
"WA treated, drinking water" = "#FCF0C0",
"WA street gutter" = "#ffe433",
"WA street gutter, sediment" = "#A03DBD",
"WA in-patient feces" = "#588c7e",
"WA treated, hand washing" = "#f4a688",
"WA river, drinking water" = "#00ced1",
"North Eu hospital effluent" = "#A1BD14"))
rownames(heat.df) <- c("aadA2_1_clust__(aadA)", "aph(6)-Id_1_clust__(aph(6)-Id)", "blaNDM-18_1_clust__(blaNDM-1)",
"blaOXA-368_1_clust__(blaOXA-10)", "blaOXA-370_1_clust__(blaOXA-48)", "blaOXA-392_1_clust__(blaOXA-1)",
"blaIMP-58_1_clust__(blaIMP)", "blaCMY-150_2_clust__(blaCMY-2)", "blaCTX-M-211_1_clust__(Group I blaCTX-M)",
"blaCTX-M-4_1_clust__(Group II blaCTX-M)", "blaCTX-M-110_1_clust__(Group III blaCTX-M)",
"blaKPC-34_1_clust__(blaKPC-2)", "blaVEB-1_1_clust__(blaVEB-1)", "cfxA_1_clust__(cfxA)",
"blaVIM-48_1_clust__(blaVIM-7)", "blaGES-1_1_clust__(blaGES-5)", "mcr-1.11_1_clust__(mcr-1)",
"mcr-3.1_1_clust__(mcr-3)", "mcr-5.1_1_clust__(mcr-5)", "erm(B)_18_clust__(ermB)", "mph(E)_1_clust__(mphE)",
"erm(33)_1__(ermA)", "nimA_1__(nimA)", "nimE_1_clust__(nimE)",
"catB3_2_clust__(catB3)", "cmlA1_1_clust__(cmlA)", "qnrA1_1_clust__(qnrA)",
"sul1_11_clust__(sul1)", "tet(A)_6_clust__(tetA)", "dfrA1_1_clust__(dfrA)")
# Row annotation
metadata_row <- data.frame(temp$V2, row.names=rownames(heat.df))
colnames(metadata_row) <- c("AB class")
## Plot log
newnames <- lapply(
rownames(heat.df),
function(x) bquote(italic(.(x))))
heat <- pheatmap(log10(heat.df + 0.000001), cluster_rows = F, cluster_cols = T,
cellwidth = 9, cellheight = 17,
border_color = "white",
fontsize=10,
colorRampPalette(brewer.pal(9, "Blues"))(100), main = "",
angle_col = 90, legend = TRUE, fontsize_row = 12, fontsize_col = 9,
labels_row = as.expression(newnames),
# filename = "selected_all_heat.png",
annotation_row = metadata_row,
annotation_col = category,
annotation_colors = ann_colors,
clustering_distance_cols = "euclidean",
treeheight_col = 18,
cutree_cols = 4, gaps_row = c(2, 16, 19, 22, 24, 26, 27, 28, 29)
)
heat
## Heatmap for taxa
metaphlan_PHY_Species <- tax_glom(metaphlan_PHY_stat, taxrank = "Species")
# ESKAPE and other relevant taxa
selected <- subset_taxa(metaphlan_PHY_Species, Species == "Acinetobacter_baumannii" | Species == "Acinetobacter_nosocomialis" | Species == "Acinetobacter_towneri" | Species == "Acinetobacter_bohemicus" | Species == "Acinetobacter_parvus" | Species == "Acinetobacter_johnsonii" | Species == "Enterobacter_cloacae_complex" | Species == "Enterococcus_faecium" | Species == "Klebsiella_pneumoniae" | Species == "Staphylococcus_aureus" | Species == "Stenotrophomonas_maltophilia" | Species == "Pseudomonas_aeruginosa_group")
# OTU matrix
heat_OTU = as(otu_table(selected), "matrix")
# Coerce to data.frame
heat.df = as.data.frame(heat_OTU)
# Tax table matrix
heat_tax = as(tax_table(selected), "matrix")
# Swap colnames
match <- match(rownames(heat.df), rownames(heat_tax))
temp <- heat_tax[match,]
all(rownames(temp) == rownames(heat_OTU))## [1] TRUEall(rownames(temp) == rownames(heat_tax))## [1] TRUErownames(heat.df) <- temp[, 7]
new_df <- heat.df[ order(row.names(heat.df)), ]
new_tax = heat_tax
rownames(new_tax) <- paste(selected@tax_table[,7])
new_tax[ order(row.names(new_tax)), ]## Kingdom Phylum Class
## Acinetobacter_baumannii "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Acinetobacter_bohemicus "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Acinetobacter_johnsonii "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Acinetobacter_nosocomialis "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Acinetobacter_parvus "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Acinetobacter_towneri "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Enterobacter_cloacae_complex "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Enterococcus_faecium "Bacteria" "Firmicutes" "Bacilli"
## Klebsiella_pneumoniae "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Pseudomonas_aeruginosa_group "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Staphylococcus_aureus "Bacteria" "Firmicutes" "Bacilli"
## Stenotrophomonas_maltophilia "Bacteria" "Proteobacteria" "Gammaproteobacteria"
## Order Family
## Acinetobacter_baumannii "Pseudomonadales" "Moraxellaceae"
## Acinetobacter_bohemicus "Pseudomonadales" "Moraxellaceae"
## Acinetobacter_johnsonii "Pseudomonadales" "Moraxellaceae"
## Acinetobacter_nosocomialis "Pseudomonadales" "Moraxellaceae"
## Acinetobacter_parvus "Pseudomonadales" "Moraxellaceae"
## Acinetobacter_towneri "Pseudomonadales" "Moraxellaceae"
## Enterobacter_cloacae_complex "Enterobacterales" "Enterobacteriaceae"
## Enterococcus_faecium "Lactobacillales" "Enterococcaceae"
## Klebsiella_pneumoniae "Enterobacterales" "Enterobacteriaceae"
## Pseudomonas_aeruginosa_group "Pseudomonadales" "Pseudomonadaceae"
## Staphylococcus_aureus "Bacillales" "Staphylococcaceae"
## Stenotrophomonas_maltophilia "Xanthomonadales" "Xanthomonadaceae"
## Genus Species
## Acinetobacter_baumannii "Acinetobacter" "Acinetobacter_baumannii"
## Acinetobacter_bohemicus "Acinetobacter" "Acinetobacter_bohemicus"
## Acinetobacter_johnsonii "Acinetobacter" "Acinetobacter_johnsonii"
## Acinetobacter_nosocomialis "Acinetobacter" "Acinetobacter_nosocomialis"
## Acinetobacter_parvus "Acinetobacter" "Acinetobacter_parvus"
## Acinetobacter_towneri "Acinetobacter" "Acinetobacter_towneri"
## Enterobacter_cloacae_complex "Enterobacter" "Enterobacter_cloacae_complex"
## Enterococcus_faecium "Enterococcus" "Enterococcus_faecium"
## Klebsiella_pneumoniae "Klebsiella" "Klebsiella_pneumoniae"
## Pseudomonas_aeruginosa_group "Pseudomonas" "Pseudomonas_aeruginosa_group"
## Staphylococcus_aureus "Staphylococcus" "Staphylococcus_aureus"
## Stenotrophomonas_maltophilia "Stenotrophomonas" "Stenotrophomonas_maltophilia"# Col annotation
country <- as.matrix(sample_data(selected)[["country"]])
country <- as.factor(country)
country <- data.frame(country)
colnames(country) <- c("country")
rownames(country) <- as.matrix(colnames(otu_table(selected)))
country$country <- gsub(" ", "_", country$country)
ann_colors <- list(country = c("Benin" = "#B2182B",
"Burkina_Faso" = "#44AA99",
"Finland" = "#2166AC"))
colnames(new_df) <- gsub(pattern = "_[A-Z].*", replacement = "", colnames(new_df))
## Plot log
newnames <- lapply(
rownames(new_df),
function(x) bquote(italic(.(x))))
heat <- pheatmap(sqrt(new_df), cluster_rows = F, cluster_cols = T,
border_color = "grey",
colorRampPalette(brewer.pal(9, "Blues"))(100),
main = "Relative abundance of clinically relevant species,\n (Metaphlan3, square root transormed)",
angle_col = 90, legend = TRUE, fontsize_row = 11,
labels_row = as.expression(newnames),
filename = "eskape_heat.png",
annotation_col = country,
clustering_distance_cols = "euclidean",
show_colnames = T,
cellwidth = 13,
cellheight = 26,
gaps_row = rep(c(1, 6, 7, 8, 9, 10, 11)),
annotation_colors = ann_colors
)
heat15 most abundant ARGs / country
# Benin
resfinder_PHY_stat_Ben <- subset_samples(resfinder_PHY_stat, country == "Benin")
resfinder_PHY_stat_Ben_abun <- tax_glom(resfinder_PHY_stat_Ben, taxrank = "Gene")
# Take 15 most abundant
resfinder_PHY_stat_Ben_abun <- prune_taxa(names(sort(taxa_sums(resfinder_PHY_stat_Ben_abun), TRUE)[1:15]),
resfinder_PHY_stat_Ben_abun)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98"), 18)
rfc <- plot_bar(resfinder_PHY_stat_Ben_abun, fill = "Gene")
rfc_plot_Ben <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of ARGs (ResFinder)"))) +
ggtitle("ARG Abundance normalized to 16s rRNA") +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", face = "bold", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold"),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 20, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 24, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))
# BF
resfinder_PHY_stat_BF <- subset_samples(resfinder_PHY_stat, country == "Burkina Faso")
resfinder_PHY_stat_BF_abun <- tax_glom(resfinder_PHY_stat_BF, taxrank = "Gene")
# Take 15 most abundant
resfinder_PHY_stat_BF_abun <- prune_taxa(names(sort(taxa_sums(resfinder_PHY_stat_BF_abun), TRUE)[1:15]),
resfinder_PHY_stat_BF_abun)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98"), 18)
rfc <- plot_bar(resfinder_PHY_stat_BF_abun, fill = "Gene")
rfc_plot_BF <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of ARGs (ResFinder)"))) +
ggtitle("ARG Abundance normalized to 16s rRNA") +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", face = "bold", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold"),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 20, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 24, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))
# Finland
resfinder_PHY_stat_Fin <- subset_samples(resfinder_PHY_stat, country == "Finland")
resfinder_PHY_stat_Fin_abun <- tax_glom(resfinder_PHY_stat_Fin, taxrank = "Gene")
# Take 15 most abundant
resfinder_PHY_stat_Fin_abun <- prune_taxa(names(sort(taxa_sums(resfinder_PHY_stat_Fin_abun), TRUE)[1:15]),
resfinder_PHY_stat_Fin_abun)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98"), 18)
rfc <- plot_bar(resfinder_PHY_stat_Fin_abun, fill = "Gene")
rfc_plot_Fin <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of ARGs (ResFinder)"))) +
ggtitle("ARG Abundance normalized to 16s rRNA") +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", face = "bold", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold"),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 20, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 24, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))
# Create dataframe
Benin <- data.frame(resfinder_PHY_stat_Ben_abun@tax_table)$Gene
BF <- data.frame(resfinder_PHY_stat_BF_abun@tax_table)$Gene
Finland <- data.frame(resfinder_PHY_stat_Fin_abun@tax_table)$Gene
top_ARGs <- data.frame(Benin, BF, Finland)
## Pristine samples
resfinder_PHY_selected <- subset_samples(resfinder_PHY, alias == "BSE74" | alias == "BSE79" | alias == "BSE93" | alias == "BSE100")
resfinder_PHY_selected_abun <- tax_glom(resfinder_PHY_selected, taxrank = "Gene")
# Take 15 most abundant
resfinder_PHY_selected_abun <- prune_taxa(names(sort(taxa_sums(resfinder_PHY_selected_abun), TRUE)[1:15]),
resfinder_PHY_selected_abun)
pristine <- data.frame(resfinder_PHY_selected_abun@tax_table)$GeneInteresting samples
“clean water”
resfinder_PHY_selected <- subset_samples(resfinder_PHY, alias == "BFH27" | alias == "BFH42" | alias == "BH11" | alias == "BH13" | alias == "BH14" | alias == "BH32" | alias == "BH52" | alias == "BSE100" | alias == "BSE74" | alias == "BSE79" | alias == "BSE93" | alias == "BSE100")
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98"), 18)
rfc <- plot_bar(resfinder_PHY_selected, fill = "Class")
rfc_plot <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of ARGs (ResFinder)"))) +
ggtitle("ARG Abundance normalized to 16s rRNA") +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", face = "bold", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold"),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~category, scales = "free", space = "free", labeller = label_wrap_gen(width = 15, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 18, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))
# Taxonomy
metaphlan_PHY_selected <- subset_samples(metaphlan_PHY, alias == "BFH27" | alias == "BFH42" | alias == "BH11" | alias == "BH13" | alias == "BH14" | alias == "BH32" | alias == "BH52" | alias == "BSE100" | alias == "BSE74" | alias == "BSE79" | alias == "BSE93" | alias == "BSE100")
metaphlan_PHY_Genus <- tax_glom(metaphlan_PHY_selected, taxrank = "Genus")
metaphlan_PHY_Genus_abund <- prune_taxa(names(sort(taxa_sums(metaphlan_PHY_Genus), TRUE)[1:15]), metaphlan_PHY_Genus)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98", "#331225"), 15)
mp <- plot_bar(metaphlan_PHY_Genus_abund, fill = "Genus")
mp_plot <- mp +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of genera"))) +
ggtitle("15 Most abundant genera (Metaphlan3)") +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", face = "bold", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~category, scales = "free", space = "free", labeller = label_wrap_gen(width = 15, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 18, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))Interesting ARGs
MCR
cols <- get_palette(c("#332288", "#117733", "#52BFAD", "#88CCEE", "#DDCC77", "#F22D3D", "#FDA4B3", "#882255"), 8)
resfinder_PHY_mcr <- subset_taxa(resfinder_PHY_stat, Class == "Polymyxin")
# Other samples
#resfinder_PHY_mcr_other <- subset_samples(resfinder_PHY, alias == "BFH26" | alias == "BFH27" | alias == "BFH42" | alias == "BH11" | alias == "BH13" | alias == "BH14" | alias == "BH32" | alias == "BH52" | alias == "BSE100" | alias == "BSE74" | alias == "BSE93" | alias == "BSE79")
# save sums
resfinder_PHY_mcr_pruned <- tax_glom(resfinder_PHY_mcr, taxrank = "Cluster_name")
resfinder_PHY_mcr_pruned <- subset_taxa(resfinder_PHY_mcr_pruned, Cluster_name == "mcr-3.1_1_clust")
df <- data.frame(sample_sums(resfinder_PHY_mcr_pruned))
# plot
resfinder_PHY_mcr <- subset_samples(resfinder_PHY_mcr, sample_sums(resfinder_PHY_mcr) >= 0.0005)
resfinder_PHY_mcr <- subset_taxa(resfinder_PHY_mcr, taxa_sums(resfinder_PHY_mcr) != 0)
alias <- paste(resfinder_PHY_mcr@sam_data$alias)
hospital <- paste(resfinder_PHY_mcr@sam_data$hospital)
rfc <- plot_bar(resfinder_PHY_mcr, fill = "Cluster_name")
rfc_plot <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols, labels = c("mcr-1", "mcr-10", "mcr-3", "mcr-3.17", "mcr-4", "mcr-5", "mcr-7", "mcr-9")) +
labs(y = expression(atop(bold("ARGs to 16s rRNA"), atop("ResFinder")))) +
ggtitle("Relative abundance of ARGs") +
scale_x_discrete(breaks=levels(factor(rownames(sample_data(resfinder_PHY_mcr)))), labels=hospital, expression(bar("x"))) +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", angle = 90, face = "bold"),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 28, family = "Times", face = "italic"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(5, "char"),
panel.background = element_rect(fill = "#FFFDF9"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 30, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white")) + guides(fill=guide_legend(ncol=1))
rfc_plot
#ggsave(filename = "mcr_bar_plot.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Visualization using heatmap
#resfinder_PHY_mcr <- subset_taxa(resfinder_PHY_selected, Cluster_name == "mcr-5.1_1_clust" | Cluster_name == "mcr-1.11_1_clust" | Cluster_name == "mcr-3.17_1" | Cluster_name == "mcr-4.1_1_clust" | Cluster_name == "mcr-7.1_1" | Cluster_name == "mcr-9_1" | Cluster_name == "mcr-10_1" | Cluster_name == "mcr-8_1" | Cluster_name == "mcr-6.1_1" | Cluster_name == "mcr-2.1_1_clust")
#otu_table(resfinder_PHY_mcr)[otu_table(resfinder_PHY_mcr) < 0] <- 0
#otu_table(resfinder_PHY_mcr)[otu_table(resfinder_PHY_mcr) > 0] <- 1
#heat_df_mcr = otu_table(resfinder_PHY_mcr)
#rownames(heat_df_mcr) <- c("mcr-1.1", "mcr-1.2", "mcr-1.3", "mcr-1.4", "mcr-1.5", "mcr-1.6", "mcr-1.7", "mcr-1.8", "mcr-1.9", "mcr-1.10", "mcr-1.11", "mcr-1.12", "mcr-1.13", "mcr-1.14", "mcr-2.1", "mcr-2.2", "mcr-3.17", "mcr-4.1", "mcr-4.2", "mcr-4.3", "mcr-4.6", "mcr-4.4", "mcr-4.5", "mcr-5.1", "mcr-5.2", "mcr-6.1", "mcr-7.1", "mcr-8", "mcr-9", "mcr-10")
#colnames(heat_df_mcr) <- sample_data(resfinder_PHY_mcr)$alias
#pheatmap(heat_df_mcr, cluster_cols = F, cluster_rows = T)Interesting ARGs
MRSA & VRE & tigecycline ARGs
# mec variants
resfinder_PHY_Cluster <- subset_taxa(resfinder_PHY_stat, Cluster_name == "mecA_1_clust" | Cluster_name == "mecB_1" | Cluster_name == "mecC2_1_clust" | Cluster_name == "mecD_1")
# van variants
resfinder_PHY_Cluster <- subset_taxa(resfinder_PHY_stat, Cluster_name == "VanHAX_1_clust" | Cluster_name == "VanHAX_PA" | Cluster_name == "VanHBX_1_clust" | Cluster_name == "VanC1XY_1_clust" | Cluster_name == "VanC3XY_1_clust" | Cluster_name == "VanHDX_6" | Cluster_name == "VanHDX_7_clust" | Cluster_name == "VanHDX_3_clust" | Cluster_name == "VanHFX_1" | Cluster_name == "VanE_1_clust" | Cluster_name == "VanGXY_1" | Cluster_name == "VanG2XY_1" | Cluster_name == "VanLXY_1" | Cluster_name == "VanHMX_1" | Cluster_name == "VanNXY_1" | Cluster_name == "VanHOX_1" | Cluster_name == "vanXmurFvanWI_1" | Cluster_name == "vanXmurFvanKWI_1" | Cluster_name == "vanXmurFvanKWI_2")
# tigecycline
resfinder_PHY_Cluster <- subset_taxa(resfinder_PHY_stat, Cluster_name == "tet(X)_1_clust" | Cluster_name == "tet(X3)_1" |
Cluster_name == "tet(X)_3_clust" | Cluster_name == "tet(37)_1" |
Cluster_name == "tet(47)_1" | Cluster_name == "tet(48)_1" |
Cluster_name == "tet(49)_1" | Cluster_name == "tet(50)_1" |
Cluster_name == "tet(51)_1" | Cluster_name == "tet(52)_1" |
Cluster_name == "tet(53)_1" | Cluster_name == "tet(54)_1" |
Cluster_name == "tet(55)_1")
resfinder_PHY_Cluster <- subset_samples(resfinder_PHY_Cluster, sample_sums(resfinder_PHY_Cluster) != 0)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98", "#331225"), 43)
rfc <- plot_bar(resfinder_PHY_Cluster, fill = "Gene")
rfc_plot <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(atop(bold("ARGs to 16s rRNA"), atop("ResFinder")))) +
ggtitle("Relative abundance of ARGs") +
theme_minimal() +
theme(axis.text.x = element_text(size = 9, family = "Times", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 20, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white"))
# Check out also with the CARD data
CARD_PHY_Cluster <- subset_taxa(CARD_PHY_stat, Class == "glycylcycline")
CARD_PHY_Cluster <- subset_taxa(CARD_PHY_Cluster, Gene == "tetX" |
Gene == "Tet(X4)" | Gene == "Tet(X3)" |
Gene == "Tet(47)" | Gene == "tet(48)" |
Gene == "tet(49)" | Gene == "tet(51)" |
Gene == "tet(56)")
CARD_PHY_Cluster <- subset_samples(CARD_PHY_Cluster, sample_sums(CARD_PHY_Cluster) != 0)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98", "#331225"), 24)
card <- plot_bar(CARD_PHY_Cluster, fill = "Gene")
card_plot <- card +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(atop(bold("ARGs to 16s rRNA"), atop("CARD")))) +
ggtitle("Relative abundance of ARGs") +
theme_minimal() +
theme(axis.text.x = element_text(size = 9, family = "Times", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 20, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white"))Interesting ARGs
Carbapenemases
resfinder_PHY_Cluster_1 <- subset_taxa(resfinder_PHY_stat, Cluster_name == "blaKPC-34_1_clust"| Cluster_name == "blaNDM-18_1_clust" | Cluster_name == "blaVIM-48_1_clust" | Cluster_name == "blaIMP-1_1_clust" | Cluster_name == "blaOXA-397_1_clust")
resfinder_PHY_Cluster_2 <- subset_taxa(resfinder_PHY_stat, Gene == "blaGES-2_1_AF326355" | Gene == "blaGES-4_1_AB116723" | Gene == "blaGES-5_1_DQ236171"
| Gene == "blaGES-6_1_AY494718" | Gene == "blaGES-14_1_GU207844" | Gene == "blaGES-16_1_HM173356"
| Gene == "blaGES-18_1_JQ028729" | Gene == "blaGES-20_1_JN596280" # carbapenemase blaGES
| Gene == "blaOXA-48_1_AY236073" | Gene == "blaOXA-162_1_GU197550" | Gene == "blaOXA-181_1_CM004561"
| Gene == "blaOXA-199_1_JN704570" | Gene == "blaOXA-204_1_KP027885" | Gene == "blaOXA-232_1_JX423831"
| Gene == "blaOXA-244_1_KP659189" | Gene == "blaOXA-245_1_JX438001" | Gene == "blaOXA-247_1_JX893517"
| Gene == "blaOXA-247_1_JX893517" | Gene == "blaOXA-514_1_KU866382" | Gene == "blaOXA-515_1_KU866383"
| Gene == "blaOXA-517_1_KU878974") # blaOXA-48-like
resfinder_PHY_Cluster <- merge_phyloseq(resfinder_PHY_Cluster_1, resfinder_PHY_Cluster_2)
#resfinder_PHY_Cluster <- subset_samples(resfinder_PHY_Cluster, sample_sums(resfinder_PHY_Cluster) != 0)
cols <- get_palette(c("#332288", "#117733", "#52BFAD", "#88CCEE", "#DDCC77", "#FDA4B3", "#F22D3D", "#882255", "#5F5E98", "#E4C960", "#FD8FD9"), 11)
names <- paste(resfinder_PHY_Cluster@sam_data$hospital)
rfc <- plot_bar(resfinder_PHY_Cluster, fill = "Cluster_name")
rfc_plot <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols, labels = c("blaGES", "blaIMP", "blaKPC", "blaNDM", "blaOXA-48", "blaOXA-58", "blaVIM")) +
labs(y = expression(atop(bold("ARGs/16s rRNA"), atop("ResFinder")))) +
ggtitle("Relative abundance of ARGs in hospital WW samples") +
scale_x_discrete(breaks=levels(factor(rownames(sample_data(resfinder_PHY_Cluster)))), labels=names, expression(bar("x"))) +
theme_minimal() +
theme(axis.text.x = element_text(size = 12, family = "Times", angle = 0, face = "bold"),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 28, family = "Times", face = "italic"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(5, "char"),
panel.background = element_rect(fill = "#FFFDF9"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 30, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white")) + guides(fill=guide_legend(ncol=1))
#ggsave(filename = "carbapenemases.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)
# Check out also with the CARD data
CARD_PHY_Cluster <- subset_taxa(CARD_PHY_stat, Class == "carbapenem")
CARD_PHY_Cluster <- subset_taxa(CARD_PHY_stat, Gene == "KPC-4" | Gene == "KPC-8" | Gene == "KPC-10" | Gene == "KPC-12" | Gene == "NDM-10" | Gene == "NDM-6" | Gene == "NDM-24" | Gene == "NDM-13" | Gene == "GES-14" | Gene == "GES-15" | Gene == "GES-18" | Gene == "GES-20" | Gene == "GES-21" | Gene == "VIM-5" | Gene == "VIM-6" | Gene == "VIM-7" | Gene == "VIM-8" | Gene == "VIM-12" | Gene == "VIM-17" | Gene == "VIM-30" | Gene == "VIM-35" | Gene == "VIM-37" | Gene == "VIM-38" | Gene == "IMP-1" | Gene == "IMP-2" | Gene == "IMP-4" | Gene == "IMP-11" | Gene == "IMP-14" | Gene == "IMP-15" | Gene == "IMP-28" | Gene == "IMP-29"| Gene == "IMP-31" | Gene == "IMP-33" | Gene == "IMP-26" | Gene == "IMP-42" | Gene == "IMP-48" | Gene == "IMP-51" | Gene == "OXA-2")
CARD_PHY_Cluster <- subset_samples(CARD_PHY_Cluster, sample_sums(CARD_PHY_Cluster) != 0)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98", "#331225"), 38)
card <- plot_bar(CARD_PHY_Cluster, fill = "Gene")
card_plot <- card +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(atop(bold("ARGs to 16s rRNA"), atop("CARD")))) +
ggtitle("Relative abundance of ARGs") +
theme_minimal() +
theme(axis.text.x = element_text(size = 9, family = "Times", angle = 45, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 20, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white"))Taxonomy
# 15 most abundant taxa in hospital WW in each country
metaphlan_PHY_Ben <- subset_samples(metaphlan_PHY_stat, country == "Benin")
metaphlan_PHY_BF <- subset_samples(metaphlan_PHY_stat, country == "Burkina Faso")
metaphlan_PHY_Fin <- subset_samples(metaphlan_PHY_stat, country == "Finland")
# At genus level
metaphlan_PHY_Genus <- tax_glom(metaphlan_PHY_Ben, taxrank = "Genus")
metaphlan_PHY_Genus_abund <- prune_taxa(names(sort(taxa_sums(metaphlan_PHY_Genus), TRUE)[1:15]), metaphlan_PHY_Genus)
#tax_table(metaphlan_PHY_Genus_abund)
# At species level
metaphlan_PHY_Species <- tax_glom(metaphlan_PHY_Ben, taxrank = "Species")
metaphlan_PHY_Species_abund <- prune_taxa(names(sort(taxa_sums(metaphlan_PHY_Species), TRUE)[1:15]), metaphlan_PHY_Species)
#tax_table(metaphlan_PHY_Species_abund)
# Check out the prevalence for some clinically relevant taxa
# Stenotrophomonas_maltophilia & Acinetobacter spp. & Enterococcus & Staphylococcus
metaphlan_PHY_Species <- tax_glom(metaphlan_PHY_stat, taxrank = "Species")
metaphlan_PHY_selected <- subset_taxa(metaphlan_PHY_Species, Species == "Stenotrophomonas_maltophilia")
metaphlan_PHY_selected <- subset_samples(metaphlan_PHY_selected, sample_sums(metaphlan_PHY_selected) != 0)
metaphlan_PHY_Genus <- tax_glom(metaphlan_PHY_stat, taxrank = "Genus")
metaphlan_PHY_selected <- subset_taxa(metaphlan_PHY_Species, Genus == "Enterococcus")
metaphlan_PHY_selected <- subset_samples(metaphlan_PHY_selected, sample_sums(metaphlan_PHY_selected) != 0)
metaphlan_PHY_Genus <- tax_glom(metaphlan_PHY_stat, taxrank = "Genus")
metaphlan_PHY_selected <- subset_taxa(metaphlan_PHY_Genus, Genus == "Staphylococcus")
metaphlan_PHY_selected <- subset_samples(metaphlan_PHY_selected, sample_sums(metaphlan_PHY_selected) != 0)
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98", "#331225"), 15)
mp <- plot_bar(metaphlan_PHY_selected, fill = "Genus")
mp_plot <- mp +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance of genera"))) +
ggtitle("Species possibly showing intrinsic resistance to carbapenems (Metaphlan3)") +
theme_minimal() +
theme(
#axis.text.x = element_blank(),
axis.text.x = element_text(size = 9, family = "Times", angle = 90, hjust = 1),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 19, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFDF9"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~country, scales = "free", space = "free", labeller = label_wrap_gen(width = 30, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 26, family = "Times", hjust = 0, vjust = 0.5, angle = 0, face = "bold"),
strip.background = element_rect(colour = "white")) + guides(fill=guide_legend(ncol=2))Global sewage, mcr (Hendriksen et al.)
GlobalSewage_ResFinder <- readRDS("GlobalSewage_ResFinder.rds")
tax_table_GlobalSewage <- data.frame(tax_table(GlobalSewage_ResFinder))
OTU_table_GlobalSewage <- data.frame(t(otu_table(GlobalSewage_ResFinder)))
GlobalSewage_metadata <- data.frame(GlobalSewage_ResFinder@sam_data)
# Add country as column to metadata (GlobalSewage_metadata)
GlobalSewage_metadata$country <- rownames(GlobalSewage_metadata)
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "Viet Nam"] <- "Viet.Nam"
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "Sri Lanka"] <- "Sri.Lanka"
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "South Africa"] <- "South.Africa"
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "New Zealand"] <- "New.Zealand"
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "Czech Republic"] <- "Czech.Republic"
rownames(GlobalSewage_metadata)[rownames(GlobalSewage_metadata) == "Cote d'Ivoire"] <- "Cote.d.Ivoire"
GlobalSewage_metadata$continent <- c("Europe", "Oceania", "Europe", "Africa", "South America", "Europe", "Africa", "North America", "Africa", "Asia", "South America", "Africa", "Europe", "Europe", "Europe", "South America", "Africa", "Europe", "Africa", "Europe", "Europe", "Africa", "Europe", "Europe", "Asia", "Middle East", "Europe", "Middle East", "Europe", "Asia", "Africa", "Europe", "Europe", "Europe", "Europe", "Asia", "Europe", "Europe", "Asia", "Europe", "Oceania", "Africa", "Europe", "Asia", "South America", "Europe", "Africa", "Europe", "Asia", "Europe", "Europe", "Africa", "Europe", "Asia", "Europe", "Europe", "Africa", "Africa", "Middle East", "North America", "Asia", "Africa")
# Check
identical(rownames(OTU_table_GlobalSewage), rownames(tax_table_GlobalSewage))## [1] TRUEall(colnames(OTU_table_GlobalSewage) == rownames(GlobalSewage_metadata))## [1] TRUEGlobalSewage_ResFinder <- phyloseq(otu_table(OTU_table_GlobalSewage, taxa_are_rows=TRUE),
tax_table(as.matrix(tax_table_GlobalSewage)), sample_data(GlobalSewage_metadata))
# All
#GlobalSewage_ResFinder_mcr <- subset_taxa(GlobalSewage_ResFinder, Gene == "mcr-3.20" | Gene == "mcr-3.4" | Gene == "mcr-1.3" | Gene == "mcr-3.8" | Gene == "mcr-1.14" | Gene == "mcr-4.3" | Gene == "mcr-3.24" | Gene == "mcr-1.13" | Gene == "mcr-3.19" | Gene == "mcr-3.15" | Gene == "mcr-8" | Gene == "mcr-3.17" | Gene == "mcr-3.11" | Gene == "mcr-3.16" | Gene == "mcr-1.2" | Gene == "mcr-1.1" | Gene == "mcr-3.6" | Gene == "mcr-4.6" | Gene == "mcr-3.21" | Gene == "mcr-3.9" | Gene == "mcr-3.25" | Gene == "mcr-2.2" | Gene == "mcr-3.2" | Gene == "mcr-3.1" | Gene == "mcr-4.4" | Gene == "mcr-3.13" | Gene == "mcr-4.2" | Gene == "mcr-1.5" | Gene == "mcr-3.12" | Gene == "mcr-3.10" | Gene == "mcr-3.23" | Gene == "mcr-4.5" | Gene == "mcr-3.3" | Gene == "mcr-3.18" | Gene == "mcr-3.5" | Gene == "mcr-4.1" | Gene == "mcr-7.1" | Gene == "mcr-3.7" | Gene == "mcr-3.14" | Gene == "mcr-5.2" | Gene == "mcr-5.1")
# Variants
GlobalSewage_ResFinder_mcr <- subset_taxa(GlobalSewage_ResFinder, Gene == "mcr-5.1" | Gene == "mcr-5.2")
cols <- get_palette(c("#332288", "#117733", "#44AA99", "#88CCEE", "#DDCC77", "#CC6677", "#AA4499", "#882255", "#5F5E98"), 41)
rfc <- plot_bar(GlobalSewage_ResFinder_mcr, fill = "Gene")
rfc_plot <- rfc +
geom_bar(stat="identity", color = NA, size = 0) +
scale_fill_manual(values = cols) +
labs(y = expression(bold("Relative abundance (ResFinder)"))) +
ggtitle("ARG Abundance normalized to 16s rRNA") +
theme_minimal() +
theme(axis.text.x = element_text(size = 10, family = "Times", face = "bold", angle = 90),
axis.text.y = element_text(size = 28, family = "Times", face = "bold", angle = 90),
axis.title.y = element_text(size = 30, family = "Times"),
axis.title.x = element_blank(),
legend.text = element_text(size = 18, family = "Times"),
legend.title = element_blank(),
legend.key = element_rect(size = 2, color = "white"),
legend.key.size = unit(1, "cm"),
legend.spacing.y = unit(2.2, "char"),
panel.background = element_rect(fill = "#FFFCF3"),
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
plot.title = element_text(size = 30, family = "Times", face = "bold")) +
facet_grid(~continent, scales = "free", space = "free", labeller = label_wrap_gen(width = 15, multi_line = TRUE)) +
theme(strip.text.x= element_text(size = 24, family = "Times", hjust = 0.5, vjust = 0.5, angle = 0),
strip.background = element_rect(colour = "white"))Mantel tests
# horn-morisita
ARG_dist <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat))), method = "horn")
MGE_dist <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat))), method = "horn")
intI1_dist <- vegdist(t(as.matrix(otu_table(MGE_PHY_int_stat))), method = "horn")
PHY_bacteria <- subset_taxa(metaphlan_PHY_stat, Kingdom %in% c("Bacteria"))
PHY_bacteria <- subset_samples(PHY_bacteria, sample_sums(PHY_bacteria) != 0)
BACT_dist <- vegdist(t(as.matrix(otu_table(PHY_bacteria))), method = "horn")
mantel(ARG_dist, MGE_dist, method = "spearman", permutations = 9999, na.rm = TRUE)##
## Mantel statistic based on Spearman's rank correlation rho
##
## Call:
## mantel(xdis = ARG_dist, ydis = MGE_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
##
## Mantel statistic r: 0.575
## Significance: 1e-04
##
## Upper quantiles of permutations (null model):
## 90% 95% 97.5% 99%
## 0.0848 0.1105 0.1303 0.1556
## Permutation: free
## Number of permutations: 9999#Mantel statistic r: 0.575
# Significance: 1e-04
#Upper quantiles of permutations (null model):
# 90% 95% 97.5% 99%
#0.0834 0.1095 0.1315 0.1585
#Permutation: free
#Number of permutations: 9999
mantel(ARG_dist, BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)##
## Mantel statistic based on Spearman's rank correlation rho
##
## Call:
## mantel(xdis = ARG_dist, ydis = BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
##
## Mantel statistic r: 0.4672
## Significance: 1e-04
##
## Upper quantiles of permutations (null model):
## 90% 95% 97.5% 99%
## 0.0931 0.1225 0.1463 0.1722
## Permutation: free
## Number of permutations: 9999#Mantel statistic r: 0.4672
# Significance: 1e-04
#Upper quantiles of permutations (null model):
# 90% 95% 97.5% 99%
#0.0927 0.1193 0.1440 0.1758
#Permutation: free
#Number of permutations: 9999
mantel(MGE_dist, BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)##
## Mantel statistic based on Spearman's rank correlation rho
##
## Call:
## mantel(xdis = MGE_dist, ydis = BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
##
## Mantel statistic r: 0.2811
## Significance: 1e-04
##
## Upper quantiles of permutations (null model):
## 90% 95% 97.5% 99%
## 0.0789 0.1012 0.1242 0.1470
## Permutation: free
## Number of permutations: 9999#Mantel statistic r: 0.2811
# Significance: 1e-04
#Upper quantiles of permutations (null model):
# 90% 95% 97.5% 99%
#0.0816 0.1048 0.1256 0.1483
#Permutation: free
#Number of permutations: 9999Mantel tests
For each country
# For each country
resfinder_PHY_stat_Ben <- subset_samples(resfinder_PHY_stat, country == "Benin")
resfinder_PHY_stat_BF <- subset_samples(resfinder_PHY_stat, country == "Burkina Faso")
resfinder_PHY_stat_Fin <- subset_samples(resfinder_PHY_stat, country == "Finland")
MGE_PHY_stat_Ben <- subset_samples(MGE_PHY_stat, country == "Benin")
MGE_PHY_stat_BF <- subset_samples(MGE_PHY_stat, country == "Burkina Faso")
MGE_PHY_stat_Fin <- subset_samples(MGE_PHY_stat, country == "Finland")
metaphlan_PHY_stat_Ben <- subset_samples(metaphlan_PHY_stat, country == "Benin")
metaphlan_PHY_stat_BF <- subset_samples(metaphlan_PHY_stat, country == "Burkina Faso")
metaphlan_PHY_stat_Fin <- subset_samples(metaphlan_PHY_stat, country == "Finland")
# Distance matrices
ARG_dist_Ben <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_Ben))), method = "horn")
ARG_dist_BF <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_BF))), method = "horn")
ARG_dist_Fin <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_Fin))), method = "horn")
MGE_dist_Ben <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_Ben))), method = "horn")
MGE_dist_BF <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_BF))), method = "horn")
MGE_dist_Fin <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_Fin))), method = "horn")
PHY_bacteria_Ben <- subset_taxa(metaphlan_PHY_stat_Ben, Kingdom %in% c("Bacteria"))
PHY_bacteria_Ben <- subset_samples(PHY_bacteria_Ben, sample_sums(PHY_bacteria_Ben) != 0)
PHY_bacteria_BF <- subset_taxa(metaphlan_PHY_stat_BF, Kingdom %in% c("Bacteria"))
PHY_bacteria_BF <- subset_samples(PHY_bacteria_BF, sample_sums(PHY_bacteria_BF) != 0)
PHY_bacteria_Fin <- subset_taxa(metaphlan_PHY_stat_Fin, Kingdom %in% c("Bacteria"))
PHY_bacteria_Fin <- subset_samples(PHY_bacteria_Fin, sample_sums(PHY_bacteria_Fin) != 0)
BACT_dist_Ben <- vegdist(t(as.matrix(otu_table(PHY_bacteria_Ben))), method = "horn")
BACT_dist_BF <- vegdist(t(as.matrix(otu_table(PHY_bacteria_BF))), method = "horn")
BACT_dist_Fin <- vegdist(t(as.matrix(otu_table(PHY_bacteria_Fin))), method = "horn")
# ARG-MGE
mantel(ARG_dist_Ben, MGE_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_BF, MGE_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_Fin, MGE_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)
# ARG-bacteria
mantel(ARG_dist_Ben, BACT_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_BF, BACT_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_Fin, BACT_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)
# MGE-bacteria
mantel(MGE_dist_Ben, BACT_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(MGE_dist_BF, BACT_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(MGE_dist_Fin, BACT_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)Mantel tests
For each country, equal sample sizes
ARG_dist <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_equal))), method = "horn")
MGE_dist <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_equal))), method = "horn")
PHY_bacteria <- subset_taxa(metaphlan_PHY_stat_equal, Kingdom %in% c("Bacteria"))
PHY_bacteria <- subset_samples(PHY_bacteria, sample_sums(PHY_bacteria) != 0)
BACT_dist <- vegdist(t(as.matrix(otu_table(PHY_bacteria))), method = "horn")
mantel(ARG_dist, MGE_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist, BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(MGE_dist, BACT_dist, method = "spearman", permutations = 9999, na.rm = TRUE)
# For each country
resfinder_PHY_stat_Ben <- subset_samples(resfinder_PHY_stat_equal, country == "Benin")
resfinder_PHY_stat_BF <- subset_samples(resfinder_PHY_stat_equal, country == "Burkina Faso")
resfinder_PHY_stat_Fin <- subset_samples(resfinder_PHY_stat_equal, country == "Finland")
MGE_PHY_stat_Ben <- subset_samples(MGE_PHY_stat_equal, country == "Benin")
MGE_PHY_stat_BF <- subset_samples(MGE_PHY_stat_equal, country == "Burkina Faso")
MGE_PHY_stat_Fin <- subset_samples(MGE_PHY_stat_equal, country == "Finland")
MGE_PHY_int_stat_Ben <- subset_samples(MGE_PHY_int_stat_equal, country == "Benin")
MGE_PHY_int_stat_BF <- subset_samples(MGE_PHY_int_stat_equal, country == "Burkina Faso")
MGE_PHY_int_stat_Fin <- subset_samples(MGE_PHY_int_stat_equal, country == "Finland")
metaphlan_PHY_stat_Ben <- subset_samples(metaphlan_PHY_stat_equal, country == "Benin")
metaphlan_PHY_stat_BF <- subset_samples(metaphlan_PHY_stat_equal, country == "Burkina Faso")
metaphlan_PHY_stat_Fin <- subset_samples(metaphlan_PHY_stat_equal, country == "Finland")
# Distance matrices
ARG_dist_Ben <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_Ben))), method = "horn")
ARG_dist_BF <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_BF))), method = "horn")
ARG_dist_Fin <- vegdist(t(as.matrix(otu_table(resfinder_PHY_stat_Fin))), method = "horn")
MGE_dist_Ben <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_Ben))), method = "horn")
MGE_dist_BF <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_BF))), method = "horn")
MGE_dist_Fin <- vegdist(t(as.matrix(otu_table(MGE_PHY_stat_Fin))), method = "horn")
PHY_bacteria_Ben <- subset_taxa(metaphlan_PHY_stat_Ben, Kingdom %in% c("Bacteria"))
PHY_bacteria_Ben <- subset_samples(PHY_bacteria_Ben, sample_sums(PHY_bacteria_Ben) != 0)
PHY_bacteria_BF <- subset_taxa(metaphlan_PHY_stat_BF, Kingdom %in% c("Bacteria"))
PHY_bacteria_BF <- subset_samples(PHY_bacteria_BF, sample_sums(PHY_bacteria_BF) != 0)
PHY_bacteria_Fin <- subset_taxa(metaphlan_PHY_stat_Fin, Kingdom %in% c("Bacteria"))
PHY_bacteria_Fin <- subset_samples(PHY_bacteria_Fin, sample_sums(PHY_bacteria_Fin) != 0)
BACT_dist_Ben <- vegdist(t(as.matrix(otu_table(PHY_bacteria_Ben))), method = "horn")
BACT_dist_BF <- vegdist(t(as.matrix(otu_table(PHY_bacteria_BF))), method = "horn")
BACT_dist_Fin <- vegdist(t(as.matrix(otu_table(PHY_bacteria_Fin))), method = "horn")
# ARG-MGE
mantel(ARG_dist_Ben, MGE_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_BF, MGE_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_Fin, MGE_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)
# ARG-bacteria
mantel(ARG_dist_Ben, BACT_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_BF, BACT_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(ARG_dist_Fin, BACT_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)
# MGE-bacteria
mantel(MGE_dist_Ben, BACT_dist_Ben, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(MGE_dist_BF, BACT_dist_BF, method = "spearman", permutations = 9999, na.rm = TRUE)
mantel(MGE_dist_Fin, BACT_dist_Fin, method = "spearman", permutations = 9999, na.rm = TRUE)Correlation of intI and all ARGs?
## Benin
# ARG names
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
args <- subset_samples(resfinder_PHY_stat, country %in% c("Benin"))
int <- subset_samples(MGE_PHY_int_stat, country %in% c("Benin"))
arg_matrix <- otu_table(args)
arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
match <- match(rownames(arg_matrix), rownames(tax))
arg_tax <- tax[match,]
rownames(arg_matrix) <- arg_tax$Gene
int_matrix <- data.frame(sample_sums(otu_table(int)))
arg_matrix <- t(arg_matrix)
Benin_correl<-corr.test(arg_matrix, int_matrix, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(Benin_correl$r)
p <- data.frame(Benin_correl$p)
p.ad <- data.frame(Benin_correl$p.adj)
Benin_cor_data <- data.frame(r, p, p.ad)
Benin_cor_data$Gene <- rownames(Benin_cor_data)
colnames(Benin_cor_data) <- c("r", "p", "p.ad", "Gene")
Benin_cor_data_filt <- Benin_cor_data[which(Benin_cor_data$p.ad < 0.05), ]
Benin_pos <- Benin_cor_data_filt[which(Benin_cor_data_filt$r > 0), ]
Benin_neg <- Benin_cor_data_filt[which(Benin_cor_data_filt$r < 0), ]
## Burkina Faso
# ARG names
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
args <- subset_samples(resfinder_PHY_stat, country %in% c("Burkina Faso"))
int <- subset_samples(MGE_PHY_int_stat, country %in% c("Burkina Faso"))
arg_matrix <- otu_table(args)
arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
match <- match(rownames(arg_matrix), rownames(tax))
arg_tax <- tax[match,]
rownames(arg_matrix) <- arg_tax$Gene
int_matrix <- data.frame(sample_sums(otu_table(int)))
arg_matrix <- t(arg_matrix)
BF_correl<-corr.test(arg_matrix, int_matrix, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(BF_correl$r)
p <- data.frame(BF_correl$p)
p.ad <- data.frame(BF_correl$p.adj)
BF_cor_data <- data.frame(r, p, p.ad)
BF_cor_data$Gene <- rownames(BF_cor_data)
colnames(BF_cor_data) <- c("r", "p", "p.ad", "Gene")
BF_cor_data_filt <- BF_cor_data[which(BF_cor_data$p.ad < 0.05), ]
BF_pos <- BF_cor_data_filt[which(BF_cor_data_filt$r > 0), ]
BF_neg <- BF_cor_data_filt[which(BF_cor_data_filt$r < 0), ]
## Finland
# ARG names
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
args <- subset_samples(resfinder_PHY_stat, country %in% c("Finland"))
int <- subset_samples(MGE_PHY_int_stat, country %in% c("Finland"))
arg_matrix <- otu_table(args)
arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
match <- match(rownames(arg_matrix), rownames(tax))
arg_tax <- tax[match,]
rownames(arg_matrix) <- arg_tax$Gene
int_matrix <- data.frame(sample_sums(otu_table(int)))
arg_matrix <- t(arg_matrix)
Finland_correl<-corr.test(arg_matrix, int_matrix, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(Finland_correl$r)
p <- data.frame(Finland_correl$p)
p.ad <- data.frame(Finland_correl$p.adj)
Finland_cor_data <- data.frame(r, p, p.ad)
Finland_cor_data$Gene <- rownames(Finland_cor_data)
colnames(Finland_cor_data) <- c("r", "p", "p.ad", "Gene")
Finland_cor_data_filt <- Finland_cor_data[which(Finland_cor_data$p.ad < 0.05), ]
Finland_pos <- Finland_cor_data_filt[which(Finland_cor_data_filt$r > 0), ]
Finland_neg <- Finland_cor_data_filt[which(Finland_cor_data_filt$r < 0), ]
## All
# intI1
# ARG names
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
args = resfinder_PHY_stat
int = MGE_PHY_int_stat
arg_matrix <- otu_table(args)
arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
match <- match(rownames(arg_matrix), rownames(tax))
arg_tax <- tax[match,]
rownames(arg_matrix) <- arg_tax$Gene
int_matrix <- data.frame(sample_sums(otu_table(int)))
arg_matrix <- t(arg_matrix)
correl<-corr.test(arg_matrix, int_matrix, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(correl$r)
p <- data.frame(correl$p)
p.ad <- data.frame(correl$p.adj)
cor_data <- data.frame(r, p, p.ad)
cor_data$Gene <- rownames(cor_data)
colnames(cor_data) <- c("r", "p", "p.ad", "Gene")
cor_data_filt <- cor_data[which(cor_data$p.ad < 0.05), ]
pos <- cor_data_filt[which(cor_data_filt$r > 0), ]
neg <- cor_data_filt[which(cor_data_filt$r < 0), ]
#write.table(neg, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/neg.txt",
# row.names=F, sep = "\t", col.names = T)
## All
# qacEdelta
# ARG names
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
args = resfinder_PHY_stat
qacEdelta <- tax_glom(MGE_PHY_stat, taxrank = "Class")
qacEdelta <- subset_taxa(qacEdelta, Class == "qacEdelta")
arg_matrix <- otu_table(args)
arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
match <- match(rownames(arg_matrix), rownames(tax))
arg_tax <- tax[match,]
rownames(arg_matrix) <- arg_tax$Gene
qacEdelta_matrix <- data.frame(sample_sums(otu_table(qacEdelta)))
arg_matrix <- t(arg_matrix)
correl<-corr.test(arg_matrix, qacEdelta_matrix, use="pairwise", method="pearson",
adjust="fdr",alpha=.05,ci=TRUE)
r <- data.frame(correl$r)
p <- data.frame(correl$p)
p.ad <- data.frame(correl$p.adj)
cor_data <- data.frame(r, p, p.ad)
cor_data$Gene <- rownames(cor_data)
colnames(cor_data) <- c("r", "p", "p.ad", "Gene")
cor_data_filt <- cor_data[which(cor_data$p.ad < 0.05), ]
pos <- cor_data_filt[which(cor_data_filt$r > 0), ]
neg <- cor_data_filt[which(cor_data_filt$r < 0), ]
#write.table(pos, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/pos.txt",
# row.names=F, sep = "\t", col.names = T)Which taxa correlate with ARG sum (draft, takes forever to run)
# Benin
#args <- subset_samples(resfinder_PHY_stat, country %in% c("Benin"))
#mp <- subset_samples(metaphlan_PHY_stat, country %in% c("Benin"))
#mp_genus <- tax_glom(mp, taxrank = "Genus")
#mp <- prune_taxa(names(sort(taxa_sums(mp_genus), TRUE)[1:12]), mp_genus)
#mp_genus
#arg_matrix <- otu_table(args)
#arg_matrix = arg_matrix[ rowSums(arg_matrix)!=0, ]
# ARG names
#tax <- data.frame(clusters_tax_table_resfinder)
#tax$sp <- rep("sp", times = 3104)
#tax$n <- rep(1:3104, each=1)
#colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
#rownames(tax) <- paste(tax$sp, tax$n, sep="")
#tax <- tax[c(-4, -5)]
#match <- match(rownames(arg_matrix), rownames(tax))
#arg_tax <- tax[match,]
#rownames(arg_matrix) <- arg_tax$Gene
#arg_matrix <- t(arg_matrix)
#metaphlan_PHY_stat_Genus_Ben <- tax_glom(mp, taxrank = "Genus")
#otu_matrix <- as.matrix(otu_table(mp_genus))
#otu_matrix = otu_matrix[ rowSums(otu_matrix)!=0, ]
#otu_matrix <- t(otu_matrix)
#Benin_correl_mp<-corr.test(arg_matrix, otu_matrix, use="pairwise", method="spearman", adjust="fdr",alpha=.05,ci=F)
#r <- data.frame(Benin_correl$r)
#p <- data.frame(Benin_correl$p)
#p.ad <- data.frame(Benin_correl$p.adj)
#Benin_cor_data <- data.frame(r, p, p.ad)
#Benin_cor_data$Gene <- rownames(Benin_cor_data)
#colnames(Benin_cor_data) <- c("r", "p", "p.ad", "Gene")
#Benin_cor_data_filt <- Benin_cor_data[which(Benin_cor_data$p.ad < 0.05), ]
#Benin_pos <- Benin_cor_data_filt[which(Benin_cor_data_filt$r > 0), ]
#Benin_neg <- Benin_cor_data_filt[which(Benin_cor_data_filt$r < 0), ]Figures for correlation matrix for ARGs & intI1
intI1 <- data.frame(sample_sums(MGE_PHY_int_stat))
colnames(intI1) <- c("intI1")
qacEdelta <- data.frame(sample_sums(MGE_PHY_qac))
colnames(qacEdelta) <- c("qacEdelta")
# DESeq2: Fin-Ben
# Benin
BenFin20 <- Ben_Fin[1:20,]
pattern_Ben_Fin <- as.matrix(BenFin20$Row.names)
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_Ben_Fin, ]
all(rownames(arg_data) == BenFin20$Row.names)## [1] TRUErownames(arg_data) <- BenFin20$Gene
# shorten gene names
rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
par(family="Times New Roman", cex=1.5)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M1 <- M[ , -c(1:20)]
p_mat1 <- p_mat[ , -c(1:20)]
# Finland
FinBen20 <- Fin_Ben[1:20,]
pattern_Fin_Ben <- as.matrix(FinBen20$Row.names)
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_Fin_Ben, ]
all(rownames(arg_data) == FinBen20$Row.names)## [1] TRUErownames(arg_data) <- FinBen20$Gene
# shorten gene names
rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
par(family="Times New Roman", cex=1.5)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M2 <- M[ , -c(1:20)]
p_mat2 <- p_mat[ , -c(1:20)]
# DESeq2: Fin-BF
# BF
BFFin20 <- BF_Fin[1:20,]
pattern_BF_Fin <- as.matrix(BFFin20$Row.names)
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_BF_Fin, ]
all(rownames(arg_data) == BFFin20$Row.names)## [1] TRUErownames(arg_data) <- BFFin20$Gene
# shorten gene names
rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M3 <- M[ , -c(1:20)]
p_mat3 <- p_mat[ , -c(1:20)]
# Finland
FinBF20 <- Fin_BF[1:20,]
pattern_Fin_BF <- as.matrix(FinBF20$Row.names)
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_Fin_BF, ]
all(rownames(arg_data) == FinBF20$Row.names)## [1] TRUErownames(arg_data) <- FinBF20$Gene
# shorten gene names
rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M4<- M[ , -c(1:20)]
p_mat4 <- p_mat[ , -c(1:20)]
# Plot with ggcorrplot
# For the legend
p_mat1[is.na(p_mat1)] = 0
p_mat2[is.na(p_mat2)] = 0
p_mat3[is.na(p_mat3)] = 0
p_mat4[is.na(p_mat4)] = 0
m0 <- ggcorrplot(M1, p.mat = p_mat1, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times", size = 9, angle = 20),
plot.title = element_text(size=9, face="bold", family = "Times"),
legend.title = element_blank(),
legend.text = element_text(family = "Times", size = 20),
legend.key.size = unit(1.4, "cm")))
title1 <- ggdraw() + draw_label("ARGs enriched in comparison Benin-Finland",
fontface = 'bold', x = 0.32, hjust = 0.1, y = 0.35, fontfamily = "Times", size = 26)
title2 <- ggdraw() + draw_label("ARGs enriched in comparison BF-Finland",
fontface = 'bold', x = 0.35, hjust = 0.1, y = 0.35, fontfamily = "Times", size = 26)
m1 <- ggcorrplot(M1, p.mat = p_mat1, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none", plot.margin = unit(c(0, 0, 0, 0), "cm"),
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 20),
axis.text.x.bottom = element_text(size = 16, angle = 35, face = "italic"),
plot.title = element_text(size=26, face="bold", family = "Times"))) + ggtitle("Benin")
m2 <- ggcorrplot(M2, p.mat = p_mat2, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none", plot.margin = unit(c(0, 0, 0, 0), "cm"),
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 20),
axis.text.x.bottom = element_text(size = 16, angle = 35, face = "italic"),
plot.title = element_text(size=26, face="bold", family = "Times"))) + ggtitle("Finland")
m3 <- ggcorrplot(M3, p.mat = p_mat3, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none", plot.margin = unit(c(0, 0, 0, 0), "cm"),
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 20),
axis.text.x.bottom = element_text(size = 16, angle = 35, face = "italic"),
plot.title = element_text(size=26, face="bold", family = "Times"))) + ggtitle("Burkina Faso")
m4 <- ggcorrplot(M4, p.mat = p_mat4, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none", plot.margin = unit(c(0, 0, 0, 0), "cm"),
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 20),
axis.text.x.bottom = element_text(size = 16, angle = 35, face = "italic"),
plot.title = element_text(size=26, face="bold", family = "Times"))) + ggtitle("Finland")
# Extract the legend from one of the plots
legend <- get_legend(m0)
# Some inception with cowplot...
A <- plot_grid(title1, m1,m2, NULL, ncol = 1, rel_heights = c(0.5, 1, 1, 0.1))
B <- plot_grid(NULL, title2, m3,m4, ncol = 1, rel_heights = c(0.1, 0.5, 1, 1))
AB <- plot_grid(A, B, ncol = 1)
#plot_grid(AB, legend, ncols = 2, rel_widths = c(2, 1))
AB
#ggsave(filename = "ARGs_corr_deseq.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Same but for 15 most abundant ARGs
#top_ARGs
MGE_PHY_qac <- tax_glom(MGE_PHY_stat, taxrank = "Class")
MGE_PHY_qac <- subset_taxa(MGE_PHY_qac, Class == "qacEdelta")
intI1 <- data.frame(sample_sums(MGE_PHY_int_stat))
colnames(intI1) <- c("intI1")
qacEdelta <- data.frame(sample_sums(MGE_PHY_qac))
colnames(qacEdelta) <- c("qacEdelta")
# Benin
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
tax$Row.names <- paste(tax$sp, tax$n, sep="")
pattern_Ben <- data.frame(dplyr::filter(tax,
grepl("aph\\(6\\)-Id_1_M28829|aph\\(6\\)-Id_4_CP000971|aph\\(3''\\)-Ib_3_AF321550|ant\\(3''\\)-Ia_1_X02340|aph\\(3''\\)-Ib_2_AF024602|aph\\(3''\\)-Ib_5_AF321551|aph\\(6\\)-Id_2_AF024602|blaOXA-347_1_ACWG01000053|blaOXA-129_1_FJWZ01000025|blaOXA-256_1_HE616889|cmlA1_1_M64556|sul1_5_EU780013|tet\\(A\\)_6_AF534183|tet\\(C\\)_2_AY046276|tet\\(C\\)_3_AF055345",
Gene)))
pattern_Benin <- as.matrix(pattern_Ben[,6])
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_Benin, ]
all(rownames(arg_data) == pattern_Benin)## [1] TRUErownames(arg_data) <- pattern_Ben$Gene
rownames(arg_data)## [1] "aph(6)-Id_1_M28829" "aph(6)-Id_4_CP000971"
## [3] "aph(3'')-Ib_3_AF321550" "ant(3'')-Ia_1_X02340"
## [5] "aph(3'')-Ib_2_AF024602" "aph(3'')-Ib_5_AF321551"
## [7] "aph(6)-Id_2_AF024602" "blaOXA-347_1_ACWG01000053"
## [9] "blaOXA-129_1_FJWZ01000025" "blaOXA-256_1_HE616889"
## [11] "cmlA1_1_M64556" "sul1_5_EU780013"
## [13] "tet(A)_6_AF534183" "tet(C)_2_AY046276"
## [15] "tet(C)_3_AF055345"# shorten gene names
#rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
rownames(arg_data) = c("aph(6)-Id_1", "aph(6)-Id_4", "aph(3'')-Ib_3", "ant(3'')-Ia_1", "aph(3'')-Ib_2", "aph(3'')-Ib_5", "aph(6)-Id_2", "blaOXA-347", "blaOXA-129", "blaOXA-256", "cmlA1", "sul1", "tet(A)_6", "tet(C)_2", "tet(C)_3")
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M1 <- M[ , -c(1:15)]
p_mat1 <- p_mat[ , -c(1:15)]
# BF
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
tax$Row.names <- paste(tax$sp, tax$n, sep="")
pattern_BF <- data.frame(dplyr::filter(tax,
grepl("aph\\(6\\)-Id_1_M28829|ant\\(3''\\)-Ia_1_X02340|aph\\(3''\\)-Ib_2_AF024602|aph\\(3''\\)-Ib_5_AF321551|blaOXA-101_1_AM412777|blaCMY-4_1_LNHZ01000079|msr\\(E\\)_1_FR751518|nimA_1_X71444|cmlA1_1_M64556|cmlA1_2_AB212941|qnrS2_1_DQ485530|sul1_5_EU780013|tet\\(39\\)_1_KT346360|tet\\(C\\)_2_AY046276|tet\\(C\\)_3_AF055345",
Gene)))
pattern_BurkinaFaso <- as.matrix(pattern_BF[,6])
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_BurkinaFaso, ]
all(rownames(arg_data) == pattern_BurkinaFaso)## [1] TRUErownames(arg_data) <- pattern_BF$Gene
rownames(arg_data)## [1] "aph(6)-Id_1_M28829" "ant(3'')-Ia_1_X02340"
## [3] "aph(3'')-Ib_2_AF024602" "aph(3'')-Ib_5_AF321551"
## [5] "blaOXA-101_1_AM412777" "blaCMY-4_1_LNHZ01000079"
## [7] "msr(E)_1_FR751518" "nimA_1_X71444"
## [9] "cmlA1_1_M64556" "cmlA1_2_AB212941"
## [11] "qnrS2_1_DQ485530" "sul1_5_EU780013"
## [13] "tet(39)_1_KT346360" "tet(C)_2_AY046276"
## [15] "tet(C)_3_AF055345"# shorten gene names
#rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
rownames(arg_data) = c("aph(6)-Id_1", "ant(3'')-Ia_1", "aph(3'')-Ib_2", "aph(3'')-Ib_5", "blaOXA-101", "blaCMY-4", "msr(E)", "nimA", "cmlA1_1", "cmlA1_2", "qnrS2", "sul1", "tet(39)", "tet(C)_2", "tet(C)_3")
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M2 <- M[ , -c(1:15)]
p_mat2 <- p_mat[ , -c(1:15)]
# Finland
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
tax$Row.names <- paste(tax$sp, tax$n, sep="")
pattern_Fin <- data.frame(dplyr::filter(tax,
grepl("aac\\(6'\\)-Ib-Hangzhou_1_FJ503047|aph\\(6\\)-Id_1_M28829|aac\\(6'\\)-IIa_1_M29695|ant\\(3''\\)-Ia_1_X02340|aadA1b_1_M95287|blaOXA-427_1_KX827604|blaMOX-6_1_GQ152601|cfxA6_1_GQ342996|blaOXA-296_1_APOH01000009|erm\\(B\\)_18_X66468|msr\\(E\\)_1_FR751518|mph\\(E\\)_1_DQ839391|mph\\(E\\)_2_JF769133|tet\\(39\\)_1_KT346360|tet\\(Q\\)_1_L33696",
Gene)))
pattern_Finland <- as.matrix(pattern_Fin[,6])
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_Finland, ]
all(rownames(arg_data) == pattern_Finland)## [1] TRUErownames(arg_data) <- pattern_Fin$Gene
rownames(arg_data)## [1] "aac(6')-Ib-Hangzhou_1_FJ503047" "aph(6)-Id_1_M28829"
## [3] "aac(6')-IIa_1_M29695" "ant(3'')-Ia_1_X02340"
## [5] "aadA1b_1_M95287" "blaOXA-427_1_KX827604"
## [7] "blaMOX-6_1_GQ152601" "cfxA6_1_GQ342996"
## [9] "blaOXA-296_1_APOH01000009" "erm(B)_18_X66468"
## [11] "msr(E)_1_FR751518" "mph(E)_1_DQ839391"
## [13] "mph(E)_2_JF769133" "tet(39)_1_KT346360"
## [15] "tet(Q)_1_L33696"# shorten gene names
#rownames(arg_data) <- gsub(pattern = "_[A-Z].*", replacement = "", rownames(arg_data))
rownames(arg_data) = c("aac(6')-Ib-Hangzhou", "aph(6)-Id_1", "aac(6')-IIa_1", "ant(3'')-Ia_1", "aadA1b", "blaOXA-427", "blaMOX-6", "cfxA6", "blaOXA-296", "erm(B)", "msr(E)", "mph(E)_1", "mph(E)_2", "tet(39)", "tet(Q)")
arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M3 <- M[ , -c(1:15)]
p_mat3 <- p_mat[ , -c(1:15)]
# Plot with ggcorrplot
# For the legend
p_mat1[is.na(p_mat1)] = 0
p_mat2[is.na(p_mat2)] = 0
p_mat3[is.na(p_mat3)] = 0
m0 <- ggcorrplot(M1, p.mat = p_mat1, type = "lower", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times", size = 30, angle = 20),
plot.title = element_text(size=14, face="bold", family = "Times"),
legend.title = element_blank(),
legend.text = element_text(family = "Times", size = 20),
legend.key.size = unit(1.4, "cm")))
m1 <- ggcorrplot(M1, p.mat = p_mat1, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none",
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 18),
axis.text.x.bottom = element_text(size = 18, angle = 35, face = "italic"),
plot.title = element_text(size=22, face="bold", family = "Times"))) +
ggtitle("15 most abundant ARGs in Benin")
m2 <- ggcorrplot(M2, p.mat = p_mat2, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none",
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 18),
axis.text.x.bottom = element_text(size = 18, angle = 35, face = "italic"),
plot.title = element_text(size=22, face="bold", family = "Times"))) +
ggtitle("15 most abundant ARGs in Burkina Faso")
m3 <- ggcorrplot(M3, p.mat = p_mat3, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none",
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 18),
axis.text.x.bottom = element_text(size = 18, angle = 35, face = "italic"),
plot.title = element_text(size=22, face="bold", family = "Times"))) +
ggtitle("15 most abundant ARGs in Finland")
# Extract the legend from one of the plots
legend <- get_legend(m0)
# Plot
A <- plot_grid(m1,m2, m3, ncol = 1)
plot_grid(A, legend, ncol = 2, rel_widths = c(1, 0.25))
#ggsave(filename = "top15_ARGs_corr_deseq.png",
# width = 16, height = 13, dpi = 300, units = "in", device='png', scale = 1)Same but for non-wastewater samples in West Africa
intI1 <- data.frame(sample_sums(MGE_PHY_int_stat))
colnames(intI1) <- c("intI1")
qacEdelta <- data.frame(sample_sums(MGE_PHY_qac))
colnames(qacEdelta) <- c("qacEdelta")
# Pristine
intI1 <- data.frame(sample_sums(MGE_PHY_int_stat))
colnames(intI1) <- c("intI1")
qacEdelta <- data.frame(sample_sums(MGE_PHY_qac))
colnames(qacEdelta) <- c("qacEdelta")
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
tax$Row.names <- paste(tax$sp, tax$n, sep="")
pattern_prist <- data.frame(dplyr::filter(tax,
grepl("aadA2_1_NC_010870|aph\\(6\\)-Id_1_M28829|ant\\(2''\\)-Ia_6_AJ871915|ant\\(2''\\)-Ia_9_HM367610|ant\\(2''\\)-Ia_10_HM367617|ant\\(3''\\)-Ia_1_X02340|blaOXA-129_1_FJWZ01000025|blaOXA-256_1_HE616889|blaADC-25_1_EF016355|blaPAU-1_1_MH053445|msr\\(E\\)_1_FR751518|mph\\(E\\)_1_DQ839391|mph\\(E\\)_2_JF769133|sul1_5_EU780013|tet\\(39\\)_1_KT346360",
Gene)))
pattern_pristine <- as.matrix(pattern_prist[,6])
args <- data.frame(otu_table(resfinder_PHY_stat))
arg_data <- args[pattern_pristine, ]
all(rownames(arg_data) == pattern_pristine)## [1] TRUErownames(arg_data) <- pattern_prist$Gene
rownames(arg_data)## [1] "aadA2_1_NC_010870" "aph(6)-Id_1_M28829"
## [3] "ant(2'')-Ia_6_AJ871915" "ant(2'')-Ia_9_HM367610"
## [5] "ant(2'')-Ia_10_HM367617" "ant(3'')-Ia_1_X02340"
## [7] "blaOXA-129_1_FJWZ01000025" "blaOXA-256_1_HE616889"
## [9] "blaADC-25_1_EF016355" "blaPAU-1_1_MH053445"
## [11] "msr(E)_1_FR751518" "mph(E)_1_DQ839391"
## [13] "mph(E)_2_JF769133" "sul1_5_EU780013"
## [15] "tet(39)_1_KT346360"arg_data = t(arg_data)
df <- data.frame(arg_data, intI1, qacEdelta)
cor <- rcorr(as.matrix(df))
M <- cor$r
p_mat <- cor$P
M5 <- M[ , -c(1:15)]
p_mat5 <- p_mat[ , -c(1:15)]
p_mat5[is.na(p_mat5)] = 0
m5 <- ggcorrplot(M5, p.mat = p_mat5, type = "full", insig = "blank", method = "square",
ggtheme = ggplot2::theme_classic() +
theme(axis.text = element_text(face = "italic", family = "Times"),
legend.position = "none",
axis.text.y.left = element_text(angle = 0, face = "bold.italic", size = 18),
axis.text.x.bottom = element_text(size = 18, angle = 35, face = "italic"),
plot.title = element_text(size=22, face="bold", family = "Times"))) +
ggtitle("15 most abundant ARGs in pristine samples")“Core” resistome and unique ARGs
Ben_temp <- otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Benin")))[rowSums(otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Benin")))) > 0]
nrow(Ben_temp) # 1738## [1] 1738BF_temp <- otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Burkina Faso")))[rowSums(otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Burkina Faso")))) > 0]
nrow(BF_temp) # 2131## [1] 2131Fin_temp <- otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Finland")))[rowSums(otu_table(subset_samples(resfinder_PHY_stat, country %in% c("Finland")))) > 0]
nrow(Fin_temp) # 1555## [1] 1555length(intersect(row.names(Ben_temp), (row.names(BF_temp)))) # 1664## [1] 1664length(intersect(row.names(BF_temp), (row.names(Fin_temp)))) # 1414## [1] 1414length(intersect(row.names(Ben_temp), (row.names(Fin_temp)))) # 1295## [1] 1295grid.newpage()
ven.p <- draw.triple.venn(area1 = nrow(Ben_temp), area2 = nrow(BF_temp), area3 = nrow(Fin_temp),
n12 = length(intersect(row.names(Ben_temp), (row.names(BF_temp)))),
n23 = length(intersect(row.names(BF_temp), (row.names(Fin_temp)))),
n13 = length(intersect(row.names(Ben_temp), (row.names(Fin_temp)))),
n123 = length(intersect(intersect(row.names(Ben_temp), (row.names(BF_temp))), row.names(Fin_temp))),
fontfamily = "Times", category = c("Benin", "Burkina Faso", "Finland"),
lty = "blank", fill = c("#B2182B", "#44AA99", "#2166AC"),
alpha = 0.75, cex = 4.5, cat.cex = 6, rotation.degree = 0, label.col = "white", cat.dist = 0.05,
filename = "Venn_diagram.png", output=TRUE, imagetype="png", margin = 0.08)
grid.draw(ven.p)
# Pairwise
grid.newpage()
p1 <- draw.pairwise.venn(area1 = nrow(Ben_temp), area2 = nrow(BF_temp),
cross.area = length(intersect(row.names(Ben_temp), (row.names(BF_temp)))),
fontfamily = "Times", category = c("Benin", "Burkina Faso"),
lty = "blank", fill = c("#B2182B", "#44AA99"),
alpha = 0.75, cex = 5, cat.cex = 6, rotation.degree = 0, label.col = "white",
cat.pos = c(300, 105), cat.dist = c(-0.07, -0.06),
ext.pos = 10,
ext.dist = -0.05,
ext.length = 0.75,
ext.line.lwd = 2,
ext.line.lty = "dashed")
grid.draw(p1)
grid.newpage()
p2 <- draw.pairwise.venn(area1 = nrow(Fin_temp), area2 = nrow(BF_temp),
cross.area = length(intersect(row.names(Fin_temp), (row.names(BF_temp)))),
fontfamily = "Times", category = c("Finland", "Burkina Faso"),
lty = "blank", fill = c("#2166AC", "#44AA99"),
alpha = 0.65, cex = 4, cat.cex = 4, rotation.degree = 0, label.col = "white", cat.dist = 0.02,
filename = "Venn_diagram.png", output=TRUE, imagetype="png", margin = 0.08)
grid.draw(p2)
grid.newpage()
p3 <- draw.pairwise.venn(area1 = nrow(Ben_temp), area2 = nrow(Fin_temp),
cross.area = length(intersect(row.names(Ben_temp), (row.names(Fin_temp)))),
fontfamily = "Times", category = c("Benin", "Finland"),
lty = "blank", fill = c("#B2182B", "#2166AC"),
alpha = 0.75, cex = 6, cat.cex = 4, rotation.degree = 0, label.col = "white", cat.dist = 0.03,
filename = "Venn_diagram.png", output=TRUE, imagetype="png", margin = 0.08, euler.d = TRUE, scaled = TRUE)
grid.draw(p3)
# And which ARGs are those?
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
match <- match(rownames(Ben_temp), rownames(tax))
Ben_names <- tax[match,]
match <- match(rownames(BF_temp), rownames(tax))
BF_names <- tax[match,]
match <- match(rownames(Fin_temp), rownames(tax))
Fin_names <- tax[match,]
#write.table(Ben_names, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/counts_Ben.txt", row.names=F, sep = "\t", col.names = T)
#write.table(BF_names, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/counts_BF.txt", row.names=F, sep = "\t", col.names = T)
#write.table(Fin_names, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/counts_Fin.txt", row.names=F, sep = "\t", col.names = T)
# What about the unique ARGs?
# Core
counts <- data.frame(otu_table(resfinder_PHY_stat))
counts[counts > 0] <- 1
core <- counts[rowSums(counts)==67,]
tax <- data.frame(clusters_tax_table_resfinder)
tax$sp <- rep("sp", times = 3104)
tax$n <- rep(1:3104, each=1)
colnames(tax) <- c("Class", "Cluster_name", "Gene", "sp", "n")
rownames(tax) <- paste(tax$sp, tax$n, sep="")
tax <- tax[c(-4, -5)]
match <- match(rownames(core), rownames(tax))
core_names <- tax[match,]
#write.table(core_names, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/core_names.txt", row.names=F, sep = "\t", col.names = T)
# Unique for Benin
temp1 <- intersect(row.names(Ben_temp), row.names(Fin_temp))
temp2 <- intersect(row.names(Ben_temp), row.names(BF_temp))
temp <- c(temp1, temp2)
temp <- data.frame(temp)
temp <- data.frame(unique(temp))
rownames(temp) <- temp$temp
unique_Ben <- data.frame(names = outersect(rownames(temp), rownames(Ben_temp)))
rownames(unique_Ben) <- unique_Ben$names
match <- match(rownames(unique_Ben), rownames(tax))
unique_Ben <- tax[match,]
#write.table(unique_Ben, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/unique_Ben.txt", row.names=F, sep = "\t", col.names = T)
# BF
temp1 <- intersect(row.names(BF_temp), row.names(Fin_temp))
temp2 <- intersect(row.names(BF_temp), row.names(Ben_temp))
temp <- c(temp1, temp2)
temp <- data.frame(temp)
temp <- data.frame(unique(temp))
rownames(temp) <- temp$temp
unique_BF <- data.frame(names = outersect(rownames(temp), rownames(BF_temp)))
rownames(unique_BF) <- unique_BF$names
match <- match(rownames(unique_BF), rownames(tax))
unique_BF <- tax[match,]
#write.table(unique_BF, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/unique_BF.txt", row.names=F, sep = "\t", col.names = T)
# Finland
temp1 <- intersect(row.names(Fin_temp), row.names(BF_temp))
temp2 <- intersect(row.names(Fin_temp), row.names(Ben_temp))
temp <- c(temp1, temp2)
temp <- data.frame(temp)
temp <- data.frame(unique(temp))
rownames(temp) <- temp$temp
unique_Fin <- data.frame(names = outersect(rownames(temp), rownames(Fin_temp)))
rownames(unique_Fin) <- unique_Fin$names
match <- match(rownames(unique_Fin), rownames(tax))
unique_Fin <- tax[match,]
#write.table(unique_Fin, "~/Documents/Metagenomes_AMRIWA/R/AMRIWA/RFiles/unique_Fin.txt", row.names=F, sep = "\t", col.names = T)